Google This is a digital copy of a book that was prcscrvod for gcncrations on library shclvcs bcforc it was carcfully scannod by Google as pari of a projcct to make the world's books discoverablc online. It has survived long enough for the Copyright to expire and the book to enter the public domain. A public domain book is one that was never subject to Copyright or whose legal Copyright term has expired. Whether a book is in the public domain may vary country to country. Public domain books are our gateways to the past, representing a wealth of history, cultuie and knowledge that's often difficult to discover. Marks, notations and other maiginalia present in the original volume will appear in this flle - a reminder of this book's long journcy from the publisher to a library and finally to you. Usage guidelines Google is proud to partner with libraries to digitize public domain materials and make them widely accessible. Public domain books belong to the public and we are merely their custodians. Nevertheless, this work is expensive, so in order to keep providing this resource, we have taken Steps to prcvcnt abuse by commcrcial parties, including placing technical restrictions on automatcd qucrying. We also ask that you: + Make non-commercial use ofthefiles We designed Google Book Search for use by individuals, and we request that you use these files for personal, non-commercial purposes. + Refrain from automated querying Do not send aulomated queries of any sort to Google's System: If you are conducting research on machinc translation, optical character recognition or other areas where access to a laige amount of text is helpful, please contact us. We encouragc the use of public domain materials for these purposes and may be able to help. + Maintain attributionTht GoogX'S "watermark" you see on each flle is essential for informingpcoplcabout this projcct andhclping them lind additional materials through Google Book Search. Please do not remove it. + Keep it legal Whatever your use, remember that you are lesponsible for ensuring that what you are doing is legal. Do not assume that just because we believe a book is in the public domain for users in the United States, that the work is also in the public domain for users in other countries. Whether a book is still in Copyright varies from country to country, and we can'l offer guidance on whether any speciflc use of any speciflc book is allowed. Please do not assume that a book's appearance in Google Book Search mcans it can bc used in any manner anywhere in the world. Copyright infringement liabili^ can be quite severe. Äbout Google Book Search Google's mission is to organizc the world's Information and to make it univcrsally accessible and uscful. Google Book Search hclps rcadcrs discover the world's books while hclping authors and publishers reach new audiences. You can search through the füll icxi of ihis book on the web at |http : //books . google . com/| \U58 m. THE NEW MADRID EARTHQUAKE U.S. GEOLOGICAL SURVEY BULLETIN 494 ^■'«i?ÄR\'5-* DEPABTHENT OF THE INTERIOB UWTED STATES GEOLOGICAL SOBVBY OBOBOB OTIB SMITH, DluciOK BOIiliVnM 49i THE NEW MADRID EARTHQÜAKE MYKON L. FÜLLER WASHINGTON GOVBBNMENT PEINTING OFFICE 1912 78 f 5Q?iii 2/99 31150-9 "- Im U.S. DEPARTMENT OF THE INTERIOR MANUEL LUJAN, Jr., Secretary U.S. GEOLOGICAL SURVEY Dallas L. Peck, Director Fiist printing 1912 Second printing 1992 Any use of trade, product, or firm names in this publication is for descriptive purposes only and does not imply endorsement by the U.S. Government Library of Congress Cataloging in Publication Data Füller, Myron L The new Madrid earthquake / by Myron L Füller. p. cm. Originally publlshed: Washington : Gov't. Print. Off.. 1912. In series: U.S. Geological Survey bulletin : 494 Indudes bibliographical references (p. ). 1. Earthquakes— Missouri— New Madrid Region. I. Title. QE535.2.U6F85 91-40833 551 .2'2'09778985-dc20 Gl P For sale by the Books and Open-File Reports Section U.S. Geological Survey, Federal Center, Box 25425, Denver, CO 80225 CONTENTS. Page. Introductioii 7 General statement 7 Field work and acknowledgments 7 The story of the earthquake 9 Souices of information 9 Summary of the disaster 10 Pievioufl earthquakes in the Miseissippi Valley 11 Recorded ediocks 11 Indian traditions 12 Geologie evidence 12 Record of the ahocks 13 Atmoepheric conditions preceding firet shock 13 Time of the shocks 13 Center of disturbance 14 Earlier shock 14 Subsequent ahocks 15 Areaaffected 16 General destructiveness of the shocks 17 Number and distribution of the shocks 17 Effects of the shocks outside of the New Madrid area 21 Mississippi Valley 21 Ohio Valley 22 Louisville 22 Cincinnati 26 Other localities 27 Great Lakes region 28 Atlantic Coastal and Piedmont plains 29 Nature of the vibrations 31 Classification of intensities 33 Periods of activity 33 Relation of distribution of shocks to time of day 35 Relation of distribution of shocks to diumal variations of barometric pres- sure 36 Relation of distribution of shocks to phases of moon 37 Relation of the earthquake and the weather 39 Direction of vibrations 40 Effect of the earthquake on life 40 Government relief 43 Phenomena of the earthquake 44 Atmospheric phenomena 44 Darkness. . : 44 OdoiB and vapors 45 "Light flashes" and "glows" 46 3 4 CONTENTS. Phenomena of the earthquake — Continued. Page. Geologie phenomena 47 Fissures 47 Record of fissuring 47 Chaiacter of the fissures 48 Types 48 Form 48 Arrangement 49 Intervals 49 Direction 49 Length 51 Depth 51 Fillings (sand dikes) 51 Objects swallowed by fissures 52 Distribution of the fissures 52 Situation 52 Localities 53 Cause of fissiuring 56 Bluff fissures 56 Fissures of the sand-blow regions 57 Fissures of the sand sloughs 57 Compound fissures 57 Faults 58 Location 58 Cause 59 Landslides 59 Chickasaw Bluffs 59 Location and character 59 Cause 61 Warping 62 Uplifts and doming 62 The records 62 Tipton ville dome 63 Blytheville dome 63 Little River dome 64 Other domes 64 Cause of uplifts and doming 64 Depression — "sunk lands " 64 Types 65 Form and trend 66 Amount of subsidence 67 Distribution 68 Cause of sinking 74 Extrusion 76 Records 76 Character of ejected material 76 Temperature of ejected water 77 Escape of gas and water after the earthquake 78 Resulting features 79 Sand blows 79 Nature 79 Situation 80 Distribution 80 CONTENTS. 5 Phenomena of the earthquake — Continued. Extrusion — Continued . Resulting features — Continued. Page. Sand sloughs 83 Nature 83 Situation 83 Distribution 84 Sand scatters 85 Cause of extrusion 85 IJndermining 87 Fault trenches 87 Sand sloughs 87 Sinks 87 Hydrologie phenomena 89 Agitation of water surfaces 89 Effect on navigation 92 Caving of banks 92 Disappearance of islands 93 Snags and sawyers 94 Floating wreckage 94 Changes in Springs 95 Action of the earthquake on forests 95 Splitting of trees 95 Destruction of forests by waves and caving banks 96 Overthrow of forests by landslides 96 General prostration of forests by vibrations 97 Dead trees 98 Uplifted trees 98 Submerged forests 98 Area of destruction 98 Effect on artificial structures 99 New Madrid region 99 Distant localities 100 Other physical phenomena 101 Noises 101 Origin and cause of the New Madrid earthquakes 102 Populär beliefs 102 Evidence of origin 103 Location of centrum 104 Ultimate cause 105 Significance of doming 105 Cause of uplift and doming 106 Contemporaneous disturbances 109 Possibilities of future disturbances 109 Bibliography of New Madrid earthquake 111 ILLUSTRATIONS. Page. Plate I. Map of earthquake features of the New Madrid district In pocket II. Af View of one side of fault trench or ''fissure" near banks of St. Franciä River, Ark.; B, Landslide trench and ridge resulting from the New Madrid earthquake, Chickasaw Bluffs, Tenn 12 III. Earthquake fissure fiUed with intruded sand, Charleston, Mo 52 IV. Aj Trees tilted by New Madrid earthquake, Chickasaw Bluffs, Tenn. ; Bf Trunks of cypress trees killed by submeigence resulting from New Madrid earthquake, Lake St. iSrancis, Ark 60 V. A, Water weeds of shallower lakes of the sunk lands, Vamey River, Mo. ; B, River-swamp type of sunk lands formed by New Madrid earthquake, Vamey River, Mo 64 VI. ^, Yoimg growth along edges of sunk lands, Vamey River, Mo. ; By Channel of Vamey River, Mo 70 VII. A, Stumps of timber killed by submeigence caused by the New Madrid earthquake, Reelfoot Lake, Tenn.; J?, Elevated cypieea near south end of Reelfoot Lake, Tenn 72 VIII. Af Sand containing lignite particles from deep well at Memphis, Tenn. ; By Sand craterlets formed near San Francisco in the earth- quake of 1906, showing the original appearance of sand blows of the tjrpe illustrated in Plate IX 76 IX. ^, Present appearance of sand blow of craterlet type in the Arkansas area of dlsturbance; By Coalescent or Imear blows obstructing drainage in the Arkansas district 80 X. Ay Stumps of trees killed by deposits of sand; J?, Trees with double sets of roots at south end of Reelfoot Lake, Tenn 84 FiGURE 1. Map showing the extent of earthquake disturbances in the New Afedrid area in 1811-12 8 2. Diagram showing earthquake activity by weeks 34 3. Diagram showing earthquake activity by intensities 35 4. Diagram showing earthquake activity by hours of the day 36 5. Diagram showing the relation of shocks to phases of the moon 38 6. Diagram showing trench or ''fissure" formed by settling of feiult blocks 48 7. Sketch map showing earthquake featiu'es southwest of Lake St. Francis 50 8. Plan of earthquake fissure at Beechwell, near Campbell, Mo 54 9. Longitudinal section along stream crossing simk lands 6& 10. Form of Channels in sunk-land district along St. Francis River 66 11. Forms of bayou Channels and bayou lakes near Mississippi River. . 67 12. Map of Reelfoot Lake and its submeiged drainage Channels 69 13. Section showing cjrpress killed by submeigence and young growth springing up in the less deeply submeiged areas 71 14. Section across submeiged Channel of Reelfoot Lake 74 15. Diagram showing normal movements of ground water, such as give rise to small Springs along streams 78 16. Profiles of sand blows and prairie moimds 79 17. Sketch showing changes in the Channel of Mississippi River at New Madrid after the earthquake of 1811-12 93 18. Diagram illustrating possible cause of doming in the New Madrid area • 108 6 THE NEW MADRID EARTHQÜAKE. By Myron L. Füller. INTRODUCTION. GENEBAIi STATEMENT. The succession of shocks designated coUectively the New Madrid earthquake occurred in an area of the central Mississippi Valley including southeastem Missouri, northeastern Arkansas, and western Kentucky and Tennessee. (See PL I and fig. 1 .) Beginning December 16, 1811, and lasting more than a year, these shocks have not been surpassed or even equaled for number, continuance of disturbance, area affected, and severity by the more recent and better-known shocks at Charieston and San Francisco. As the region was almost unsettled at that time relatively little attention was paid to the phe- nomenon, the published accounts being few in number and incom- plete in details. For these reasons, although scientific literature in this country and in Europe has given it a place among the great earthquakes of the worid, the memory of it has lapsed from the public mind. Shaler,^ writing of the earthquake in 1869, said: The occurrence of such a shock in a region like the Mississippi Valley, on the bor- ders ofa great river,is probablyunprecedented in the historyof earthquakes. * * * Many of the events of that convulsion were without a parallel. Scientifically this earthquake may be regarded as a type, exhibit- ing in unusual detail the geologic effects of great disturbances upon unconsolidated deposits. For this reason its phenomena have an importance which, in the absence of any previous systematic dis- cussion, Warrants detailed consideration. FIELD WÖBE AND ACKNOWLEDGMENTS. The writer's attention was first called to the region by Prof. E. M. Shepard, who had become interested in it because of the relation between artesian conditions and certain effects of the earthquake, which had become apparent during an investigation of Underground waters. In the fall of 1904 Prof. Shepard and the writer traversed 1 Shaler, N. S., Earthqaakes of the western United States: Atlantic Monthly, Nov., 1869, pp. 549-559. 7 8 THE KEW MADBID EABTHQTTAKB. in a dugout the sunk lands along Vamey River, near Kennett, Mo., and later made a trip on horseback up the old De Soto trail and along St. Francis River. The second trip was made under the guidance Arwof LEGEND Am«f Mittat «hieb boCh nint flt whicli flB^ FiOUBB 1.— Map ahowing the extent of earthquake cUBturbanoes in the New liadrid area In 1811-12. For additional reoords ol the diieotion of vibratJona see table, page 41. of Mr. C. B. Baily, city engineer of Wynne, Ark., who from timber expiorations had become familiär with the earthquake f eatures in the still almost untouched f orests north of the St. Louis^ Iron Mountain & SOUBCES OF INFORMATION. 9 Southern Railway between Memphis and Wynne. In the foUowing year the writer, m connection with studies of Underground water, covered most of the region reached by railroads in Arkansas and Missouri and made a wagon trip, again in Company with Prof. Shepard, around Keelfoot Lake, in Tennessee. A short account embodying Prof. Shepard's observations appeared in 1905,* and a number of preliminary notes and populär papers by the writer have been pub- hshed in magazines.^ In the present report an attempt is made to present a systematic record of some of the phenomena of this great earthquake, includ- ing both the contemporaneous and the present aspect of the resulting features. THE STORY OF THE EARTHQUAKE. SOUBCES OF INFORMATION. The Story of the earthquake is told in two ways — in the quaint, picturesque, and graphic accounts of contemporaries, and in the equally striking geographic and geologic records, which even now may be clearly read at many points in the region. For the sake of brevity only a single general account compiled f rom the early descriptions will be presented, but in the bibliography (pp. 111-115) references to original publications containing detailed narratives are given. In the discussion of the physiographic phenomena many references to reports of the old writers will also be found. The contemporary accounts are doubtless exaggerated, for calm Observation and accurate recording of an earthquake is impossible if the shocks are severe and dangerous. It is interesting to note, however, that, except a few features, such as the flashes of light (doubtless resulting from the general beüef in the volcanic origin of the disturbance), most of the reported phenomena have been verified by the recent investigations. Fortunately a number of scientists or men of education were in or near the region during the period of disturbance and have given vivid pictures of their experiences. John Bradbury, a noted English botanist, was on a fiatboat on the Mississippi only a few miles below New Madrid at the time of the shock; Audubon, our own naturalist, was traveling in Kentucky; Capt. Roosevelt was taking down the river the first steamer to navi- gate westem waters; L. Bringier, a well-known engineer and surveyor, was in the midst of the disturbance; Maj. Long's expedition was passing through the region on its way from Pittsburgh to the Rocky Mountains. Daniel Drake, at Cincinnati, and Jared Brooks, at 1 The New Madrid earthqaake: Jour. Oeology» vol. 13» pp. 45-62. * Caoses and periods of earthquakes in the New Madrid area, Missouri and Arkansas: Science, new ser., ▼oi. 21, 1905, pp. 349*350; Gomparative intensities of the New Madrid, Charleston, and San Francisco earth; qoakes: Idem, vol. 23, 1905, pp. 917-918; Our greatest earthquakes: Pop. Sei. Monthly,July, 1906, pp. 76-86- Earfhqaakes and the forest: Forestry and Irrigation, vol. 12, 1906, pp. 261-267. 10 THE WKW UATOtin EASTHQTTAKS. Louisville, kept careful records of the shoc^, and S. L. MitchUl, geologist and Congressman, collected records from all parts of the countiy. While the resulting pheiKHnena were stül fresh the r^on was visited by our own geographer, Tlmotiiy Flint, and by Sir Charles Lyell, the great Engliah geologist. Lyell's Visit was particulariy fortunate, for he has givai us a graphic description of the conditions 35 years aft«r the disturbance, when the Vegetation had not yet hidden the evidences of the movement. This description, with later observations on the geologic features, many of which are verypronounced, fumishes data of the greatest value and substantiates in almost every particular the eariy accomits. : SXJIOCABY OF ^SB DISA8TEB. The evening of December 15, 1811, in the New Madrid area was |- clear and quiet, with no unusual conditions which could be r^arded p as portending the catastrophe soon to take place. A little after 2 j o'clock on the moming of December 16, the inhabitants of the r^on -^ were suddenly awakened by the groaning, creaking, and cracking of ~ the timbers of the houses or cabins in which they were sleeping, by ^ the rattle of fumiture thrown down, and by the crash of f alling chim- Z neys. In fear and trembling they hurriedly groped their way from i their houses to escape the falling dä[>ris, and remained shivering in z the winter air until moming, the repeated shocks at intervals dnring the night keeping them from retmning to their weakened or tottering dwellings. Daylight brought little improvement to their Situation, for early in the moming another shock, preceded by a low rumbling and fuUy as severe as the first, was experienced. The ground rose and feil as earth waves, like the long, low swell of the sea, passed across its surface, tilting the trees until their branches interlocked and opening the soil in deep cracks as the surface was bent. Land- i slides swept down the steeper bluffs and hillsides; considerable areas \ were uplifted, and still larger areas simk and became covered with water emerging from below through fissures or little "craterlets'' or accumulating from the obstruction of the surface drainage. On the Mississippi great waves were creat^, which overwhelmed many boats and washed others high upon the shore, the retum current breaking off thousands of trees and carrying them out into the river. High banks caved and were precipitated into the river, sand bars and points of islands gave way, and whole islands disappeared. During December 16 and 17 shocks continued at short intervals but gradually diminished in intensity. They oceurred at longer intervals until January 23, when there was another shock, similar in intensity and destructiveness to the first. This shock was fol- lowed by about two weeks of quiescence, but on February 7 there were several alarming and destructive shocks, the last equaling or P PREVIOUS EABTHQUAKES IN THE MISSISSIPPI VALLEY. 11 surpassing any previous disturbance, and for several days the earth was in a nearly constant tremor. For f ully a year from this date small shocks occurred at intervals of a f ew days, but as there were no other destructive shocks the people gradually became accustomed to the vibrations and gave little or no further attention to them. PREVIOUS EARTHQUAKES IN THE MISSISSIPPI VAIiliEY. That the shock known as the New Madrid earthquake was not the first feit in the r^ion is shown by written records, by Indian tradi- tions, and by geologic evidence. Recorded shocks. — Of the shocks feit by the eariy settlers the best summary is supplied by Drake. Speaking of the series of 1811-12 he says:* The shocks of this protracted series are not all which this country has sustained since it has been the abode of civilized man. We have certain accounts of five others. The first was in the year 1776. Mr. John Heckewelder, then a missionary of the United Brethren, on the Muskingum River, in this State, has politely favored me with a memorandum conceming it. He does not recoUect the month, but it was in the Biunmer, and about 8 o'clock a. m. Its duration was two or three minutes. The Southwest side of the house was raised with such violence that the fumiture of the room was nearly overtumed. It was accompanied with a subterranean rumbling noise. Early in the moming the weather was fair, but previous to the shock it began to thicken in the southwest. The cattle were frightened by the shake, and the Indians continued, after it, to apprehend some great disaster, of which they conceived this to be the precursor. The second shock was in the year 1791 or 1792. I am unable to ascertain the precise time, but think it occurred in the month of April or May, about 7 o'clock in the moming. The weather was fair and mild. The jar was sufficient to agitate the fumi- ture of the house. A rumbling noise in the earth, which seemed to pass from west to east, preceded the shake. It was, I believe, generally feit through the northem and norüieastem parts of Kentucky, but whether beyond them I have not been able to leam. The third shock occurred, as I am informed by George Turner, Esq., about 3 o'clock a. m. January 8, 1795, at Easkaskia, Illinois Territory. It was also, I believe, feit in some parts of Kentucky. Its diuration he estimates at a minute and a half. Its direction was nearly west and east. A subteiranean noise attended, resembling that of many carriages driven rapidly over a pavement. A fourth shock was experienced, we are informed by Prof. Barton,^ at the Falls of Niagara, about 6 o'clock on the moming of the 26th of December, 1796. It appeared to come from the northwest and did not last more than two seconds, but was sensibly feit for 50 miles around the Falls. The fif th and only additional shock of which I have been fumished with any certain accounts, occurred in the southem neighborhood of Lake Michigan, .at 10 minutes past 2 o'clock p. m. on the 20th of August, 1804. At Fort Dearbom, on the bank of the lake, it was severe. From the report of Capt. William Whistler, it must have been a strenger throe than any experienced at this place. It was succeeded by a 1 Drake« Daniel, Natural and Statistical view or picture of Cincinnati, Cincinnati, 1815, pp. 243-244. t Philadelphia Medical and Physical Journal, vol. 1. 12 THE NEW MADRID EARTHQUAKE. Short huiricane from the lake. At Fort Wayne, lying considerably to the east-south- east, it was less violent. John Johnston, Esq., my Informant, remarks that the day at that place was clear and warm, without any unusual appearance. The general course of the earthquake was undoubtedly that of a line passing through those two forts. Indian trcidilions. — Lyell records * that the Indians of the Missis- sippi Valley had a tradition of a great earthquake which had previously devastated the same region, but he concluded from the absence of old sink holes and of dead trees that no convulsion of similar magnitude could have occurred for many centuries previous to 1811. As shown in the foUowing paragraphs he was mistaken in regard to the absence of such indications of previous shocks, for although it appears to be true that no fallen timber remained, there are many conspicuous and unquestionable geologic evidences of earlier disturbances. Geologie evidence. — ^The geologic evidence of shocks long ante- dating that of 1811 is very conclusive, as has elsewhere been pointed out by the writer.' Cracks as large as any of those of the last great disturbance have been seen with trees f uUy 200 years old grown on their bottoms and slopes (PI. II, ^) indicating early shocks of an intensity equal to if not greater than that of the last. Nor is the ac- tion apparently altogether recent, for post-Lafayette but pre-Iowan faults (antedating the deposition of the loess), and apparently being either a cause or accompaniment of earthquakes, have been observed by the writer in Crowley Ridge, and Glenn has described * sandstone dikes Alling old earthquake cracks in the Porters Creek formation of the Eocene Tertiary. Other geologic evidence leading to the same conclusion is seen in the Tiptonville, Blytheville, and Little River domes and in the occur- rence of certain sand sloughs. The Tiptonville dome is known to have antedated, in part at least, the shocks of 1811; as several writers mention that previous to this earthquake the land at New Madrid was never overflowed. This would not have been the case if it had been a part of the undisturbed flood piain. The erosion of the Blytheville and Little River domes since their uplift has been con- siderable (p. 64) and took place almost entirely before the 1811 shocks. If these domes are classed as earthquake featinres, as ap- parently they should be, from the description of the additional uplift of the Tiptonville dome which took place in 1811, it foUows that the original disturbance must have long antedated the New Madrid earth- quake. South of Lake St. Francis, as described elsewhere (p. 84), 1 Lyell, Charles, A seoond visIt to the United States of North America, London, 1840, p. 238. * Causes and periods of earthquakes in the New Madrid area, Missouri and Arkansas: Scienoe, new^ ser. vol. 21, 1005, pp. 340-^50. Onr greatest earthquakes: Pop. Sei. Monthly, July, 1006, p. 86. * Olenn, L. C, Underground waters of Tennessee and Kentucky west of Tennesaee River: Water-Supply Paper U. S. Geol. Survey, No. 164, 1006, pp. 30-31. TIME OF THE SHOCKS. 13 several sloughs exist; which have all the characteristics of sunk lands except the dead timber, and are apparently true earthquake features. The absence of dead timber, such as characteiizes the areas which sunk in 181 1, howeyer^ points to a considerably earUer origin. REGORD OF THE SHOCKS. ATMOSPHEBIC OONDITIONS PBECEDING FIBST SHOCX. There is apparently no possible relation between earthquake shocks und weather^ but it is not impossible that the variations in barometric pressure accompanying cyclonic movements of the atmos- phere may occasionally be a factor. No barometric records are available for the time of the first New Madrid shock; and the atmos- pheric pressure can only be inferred from the weather conditions. The best information available is that afforded by the accounts of S. L. Mitchill/ C. L. Latrobe,* Daniel Drake,* and John Hayward.* From their descriptions it appears that nothing occurred in any way suggestive to scientific xninds of unusual conditions, although it is fairly well established that immediately before the earthquake unu- sual warmth and a thick oppressive atmosphere with occasional rain and unseasonable thunder showers prevailed over a wide area of country. Near the immediate point of origin of the shocks, however, the weather seems to have been clear. The records of the shock as preserved in contemporary literature are meager, but the writer has been able to obtain, chiefly from the compilation of Jared Brooks,^ a list of some 250 shocks which took place in the six months following the one on December 16, 1811. The chief items of interest are given below. TIME OF THE SHOCKS. The exact time of the earthquake shocks must apparently remain in doubt, for, so far as is known, not a single exact determination was made. No seismographs were in existence, and as all of the principal shocks occurred in the middle of the night, when the con- sultation of reUable Chronometers was impracticable, nothing was available but personal clocks or watches. These were all supposed to be set by local time, but as the longitude of few of the places had been accurately determined, and in most localities no special attempt to fix the true sun time had been made, the times given in the table must be considered as only approximately correct. 1 A detaüed narrativB of the earfhqaakes which oocuned on the 16th day of Deoember, 1811: Trans. Lit. and Phüos. Sog. New York, vol. 1, 1815, pp. 281-307. < The rambler in North America, 2d ed., vol. 1, London, 1836, p. 107. » Natural and Statistical view or picture of Cincinnati, Cincinnati, 1815, pp. 239-242. 4 Natural and aboriginal history of Tennessee, etc., Nashville, 1823, p. 124. • MoMurtrie, H., Sketches of Ix>uisville and its environs, Louisville, 1819, Appendix. 14 TSE NEW MADRID KARTHQUAKB. The reported times of the three principal shocks ara giTen in the following table. Something of the general rate of trausmission maj be determined from them, but the data are too inaccurate to Warrant an attompt to constmct coseismal lines. ReporUd Uma of principal »hoda. Locality. Shockor Doo-l«.« Jaa.a.t SboDkot o. «. '■?-» ..m. 4.00 2.1» e-iB 3.00 11 2.M 2.00 8. so 0.30 2.80 < B«temd to merldian oI FbOadelphia. OBHTBB OF DISTÜBBANGB. Two lines of evidence are avaüable for deterDÜning the positiou of the Center of disturbance: (1) The completeness of the destruction of bnildings and the extent of geologic and phjsiographic changes, and (2) the recorded direction of moTements. Earlier skoek.—hi the first shock by far the greatest destruction occurred in the heavily-shaded area on figure 1, with minor damage in the lightly-flhaded areas aloi^ the rivers, The directions of Vibra- tion in the outlying towns, which are fairly accordant, also point to an origin in the heavily-shaded area or in the vicinity of New Madrid. In view of the fact that the waves at New Madrid and along the Mississippi were from the west, it seems certain that the centrum was west of the river, Shaler, on the basis of statements attributed to the Indians that "in the region between the Mississippi and the great plains forests were overthrown, rocks split asunder," etc., placed the "seismic Vertex of the New Madrid shocks much to the west of the Mississippi. ' ' ' Äs a matter of fact it L9 probable that the "great plüii3"referred to are the extensive prairies just west of Crowley Kidge. At any rate, it is now apparent that the effects of the shocks were much less in the r^on west of the ridge than in the area between the ridge and the Misässippi. > Shakc, N. S., BacthqDakw Ol AtlantloUoiitU;, Not., I CENTEB OF DISTUBBANOE. 15 The shock is believed to have proceeded from a northeast-«outhwest fault, as shown in Plate I, located approximately 15 miles west of the riyer. This is a few miles f arther west than the point given by the contemporary writers, namely, either New Madrid or Little Prairie (now Caruthersville) a few miles below, at which point the destruction was most complete. At that time the effects of the shock in the unsettled regions back from the river, in which the new position has been determined, were Uttle known. Suhsequeid shocka. — ^The few recorded directions of the shock of December 17 seem on the whole to favor a centnim in the same area. The shock of January 23 was instrumentally recorded as from the south-southeast at Cincinnati. It was, however, one of the most violent of the series at New Madrid, doing much damage, but the country to the east, where, on the basis of Cincinnati determinations, the centrum should have been, was little affected. It is therefore probable that the shock originated, as the earhet ones had, in the New Madrid area, the abnormal direction at Cincinnati being explained, perhaps, by local causes. Of the three "great shocks" of February 7 at Cincinnati, the first two were from the southwest or south-southwest — that is, from the direction of New Madrid — thesubsequent shock being from the south- southeast. Like the shock on January 23, this one may have been affected in direction by local causes, perhaps by a reflected motion from the Appalachian mountain mass. It is not impossible, how- ever, that new centers of disturbance were formed during this and other of the later shocks, for though the disturbances at New Madrid were little greater than the earUer shocks they were decidedly strenger in the outlying eastward districts, the distances to which the vibra- tions extended being considerably greater. From these or similar evidences Shaler concluded that, although during the first part of the series of shocks the center of the disturbance was west of New Madrid, ''the point of greatest frequency gradually moved eastward until it was near the mouth of the Wabash River in the Ohio Valley. Here, over a region about 20 miles in diameter, a succession of shocks occurred for more than two years, during which time only a few daya passed without bringing a distinct movement. Most of the oscil- lations were of such a slight character as not to be feit outside of this narrow district." ^ There is, however, no evidence of great disturbances in the shape of faults, sand-blows, sunk lands, or domes^ except in the New Madrid area. On the whole it seems probable, therefore, that all the severe shocks originated in the original area of disturbance, but it is not impossible that some of the smaller shocks may have had other foci, the vibrations originating in local readjustments due to the dis- t Shaler, N. S., Earthquakes of the westem United States: Atlantlo Monthly, Nov., 1860, p. 556. 16 TH£ NEW MADBID EABTHQUAKE. turbance of the equilibrium by the New Madrid shocks, either through the settling of the ioose Sediments of the Valleys or minor slipping along fault planes in the harder rocks. The vertical shocks at Detroit and Cincinnati on May 8 may thus represent local readjustments foUowing the severe general disturbance of the day before. At the United States Saline, 111., shocks were feit almost daily for two years, but in general they were seldom feit at Shawneetown, only 12 miles away. Some of the more severe shocks, however, were feit at this town as well as at Kaskaskia and along Wabash River.^ It should be emphasized, however, that although the shocks mentioned may be of local origin it is not improbable that they were the result of the dis- turbance of local geologic conditions by the New Madrid vibrations. ABEA AFFECTED. The area affected by the New Madrid earthquake may be subdi- vided into an area of marked earth disturbances, an area of slight earth disturbances, and an area of tremors only. In the first is included the territory characterized by pronounced earthquake phe- nomena, such as domes and sunk lands, fissures, sinks, sand blows, large landsUdes, etc. This district includes the New Madrid region, originally considered a relatively small area, including the villages of New Madrid and Little Prairie (Caruthersville). It is now known, however, to be somewhat larger, extending from a point west of Cairo on the north to the latitude of Memphis on the south, a dis- tance of more than 100 miles, and from Crowley Ridge on the west to Chickasaw Bluffs on the east, a distance of over 50 miles. The total area characterized by disturbances of the type mentioned is from 30,000 to 50,000 Square miles. In the area of slight earth disturbances will be included districts in which such minor features as the caving of banks, etc., took place. We have records in the narratives of Latrobe ^ and others of the occurrence of such phenomena along the Mississippi and Ohio, while Bradbury " records similar disturbances as f ar down the Mississippi as the mouth of the St. Francis, near Helena. The disappearance of Island 94 near Vicksburg has been described by August Warner.* In f act, there is little doubt that such phenomena as caving were promi- nent northward nearly to Herculaneum, northeastward to a point beyond the Wabash, and southward at least to the mouth of the Arkansas. Although no records from the White River region have been seen, it was probably included in the area of slight disturbance, and it is so shown in figure 1. It is also possible that the lower Arkansas was affected to some extent. * Drake, Daniel, Natural and Statistical view or picture of Cincinnati, Cincinnati, 1815, p. 238. * Latrobe, C. J. The rambler in North America, 2d edition, vol. 1, London, 1836« ' Bradbury, John, Early westem travels, Cleveland, 1904, vol. 5, pp. 204-210. ' Warner, August, quoted In Xm, Geologist, vol. 30, p. 83. NUMBER AND DISTRIBUTION OF THE SHOCKS. 17 The area of tremors was naturally far more extensive. On the north they are reported to have been feit m '' Upper Canada," on the northwest they are reported to have been feit by the Indians in the region of the upper portions of the Missouri country,* and in the region between the headwaters of the Arkansas and the Missouri, a distance of more than 500 miles from New Madrid. Southwestward the shocks were feit in the Red River Settlements and on the Washita River, an equal distance from the center of disturbance. To the south the shock was feit at New Orleans, also 500 miles distant ; to the northeast at Detroit, 600 miles away; and to the east at Wash- ington, over 700 miles, and at Boston, 1,100 miles distant. A total area of over 1,000,000 Square miles, or half that of the entire United States, was so disturbed that the vibrations could be feit without the aid of Instruments. GENEBAL DESTBUCTIVENESS OF THE SHOCKS. The severity of the shocks was such as to ruin the country in the central area of disturbance for years. After the earthquake had moderated, according to Flint,^ the country — exhibited a melancholy aspect of chafims, of sand coyering the earth, of trees thrown down, or lying at an angle of 45°, or split in the middle. The earthquakes still recurred at Short intervals, so that the people had no confidence to rebuild good houees, or chinineys of brick. The people of Little Prairie (CaruthersviQe), who suffered most — had their settlement — which consisted of a hundred families and which was located in a Wide and very deep and f ertile bottom — ^broken up . When I passed it, and stopped to contemplate the traces of the catastrophe which remained after seven years, the crevices where the earth had burst were sufficiently manifest, and the whole region was covered with sand to the depth of 2 or 3 feet. The surface was red with oxided pyrites of iron, and the sand blows, as they were called, were abundantly mixed with this kind of earth, and with pieces of pit coal. But two families remained of the whole settlement. * * * When I resided there, this district, formerly so level, rieh, and beautiful, had the most melancholy of all aspect of decay, the tokens of former culti- vation and habitancy, which were now mementos of desolation and desertion. Large and beautiful orchards, left uninclosed, houses uninhabited, deep chasms in the earth, obvious at frequent intervals — such was the face of the country, although the people had for years become so accustomed to frequent and small shocks, which did no essential injury, that the lands were graduaJly rising again in value, and New Madrid was slowly rebuüding, with fraü buildings, adapted to the apprehensions of the people. NiniBEB AND DISTBIBUTION OF THE SHOCKS. The number and distribution of the principal shocks can be best presented by means of a table such as is given below. It should be bome in mind that numerous as the recorded shocks are, they are « James, Edwin, Aoooiint of an expedition from Pittsbnrgh to tbe Rocky Mountains, Philadelphia, 1823, YoL 1, p. 272. > Fttnt, Timothy, Reoollectiöns of the last ten years, Boston, 1826, pp. 225, 227. 18 THE NEW MADRID EABTHQUAKE. but a small pari of the nmnber that would have been recorded if there had been any seismographs in the zone of distnrbance. Unfortu- nately, however, no instnunents existed in the region except some homemade apparatns at Cincinnati, and only those shocks stroag enough to be feit or to affect objects visibly are in general noted. Record of the shocks of the New Madrid earthquake} Authority. Bradbury. Bnran... Losieur. MitchiU.... Locality. Between New Madrid and mouth of 8t. Francis River. do do .do... .do .do... .do... .do .do... .do... .do... .do... .do... .do... .do.. .do.. .do.. .do.. .do.. .do.. New Madrid. — .do....... do do do do.... — .do.... do — .do.... Cape Oirardeau . Waäiington.... do do...... Richmond. do.... do.... Norfolk and Ports- mouth, Va. Raleigh,N.C do.... Geonetown, S. C . do.... Columbia, S.C.... .....do do — .do Laurens and New- berry. 8. C. Charleston, 8. C. . .do. Date. Dec. 16,1811 .do. .do. .do. .do. .do. .do. Dec. 17,1811 do do do Dec. 18,1811 do do • • • • • U w «••«• • • ■ do Dec. 19,1811 /Dec. 19,(18?), t 1811. Dec. 20,1811 Deo. 21,1811 I Jan. 23,1812 Jan. 23,1812 to Feb. 4,1812 Feb. 4, 1812 Feb. 5,1812 Feb. 7,1812 Dec. 16,1811 do June 7, 1812 Feb. 7, 1812 Nov. 9,1812 Dec. 16,1811 .do. .do. .do. .do. .do. .do. do.... — do — .do... do Dec. 18,1811 Dec. 16,1811 — .do. ....do Dec. 17,1811 ....do.... Dec. 16,1811 .do Sa. m Hour. 2 a. m. 2.30orafter IntervalsofetolO minutes during night; 27 shocks belore daylight. Daylight BreaküBst After breakfttst 11 a. m 5 a. m 7 a. m 12noon... 7.30 p. m.. Stola, m. 6 a. m 12noon ... } 6 p. m. 9 p. m. 11 5 p. m. 7p. m.... 4.30 a. m. I- 2 a. m. 7 a. m. 4 p. m. 3 a. m. 6 a. m. 8 a. m. 3 a. m. 6 a. m. 8 a. m. 3 a. m. 8 a. m 2 to 3 a. m 3 to 7 a. m 3 to 8 a. m 12 noon . . . 2.30 a. m . . 8 a. m 10 a. m 12.15 p. m. 12.15 p. m. 3 a. m Remarks. Violent; boat nearly upset; trees feil; bankscaved. Terrible, but not equal to flrst. Slight compared with first and sec- ond. Equal to first; same phenomena. Very severe; nearly thrown down. Man nearly tiirown into river. Violent; trees shaken; banks feil; river agitated. Severe and of long duration. Violent and of long duration; trees thrown into river. 81ight. Do. Not venr violent, but lasted nearly a minute. As violent as worst of preceding. Shocks frequent. Nearly as severe as any. 4 shocks, severe. Ejiown as hard shock. As severe as first. Big shock. Considerable motion to fumiture. Streng enough to shaks Windows and fumiture. Do. Light. Shook houses and fumiture and rangbells. Do. Do. Stopped clocks, rattled doors, swung hanging objects. Do. Several shocks distinctly feit. Several faint shocks. Severe; upset tub of water, etc. Slight. Houses rocked'; plaster feil; dogs barked; heavy shock foUowed by 3 light ones. 2 shocks. Light. Smart shock. Cracked chimneys. Lasted 2 to 3 minutes; bells rang; clocks stopped; well water roUed; buildings shaken. Do. 1 Except at Louisville, for which see more detailed table, pp. 22-26. KUMBEB AND DISTRIBUTION OF THE SH0GK8. 19 Reoord o/the thocks o/the New Madrid earthquake — Gontinued. Authority. Locality. Ifltehfll James. Brake. Cbaile8toii,S.C.. SaTannah, Oa.... do > • • • •Uw««««« • • • fl Natches. Miss. Piney River, Teim do.... do Enoxville, Tenn. . — .do.... do.... Columbia, Tenn... Jefferaonville, Ind. Vinoennes, Ind.... Red Bank, Ind. do St. Louis, Mo .do. .do. .do. • ■ ■ ••Uv* ••••••••••■ ■ • « • »laU •«•••••••••• Lebanon, Ohio.. . . Aichville, Ohio... Ilenderson Coon- ty,Ky. Detroit, Mich Heroolanemn, Mo. do do Carthage, Tenn... do Henderson, Ark... New Orleans, La. Philadelphia, Pa. New York Newark, N. J Washington, D.C. Nottingnam, Ind.. Rjchmond, Va Codiocton, Ohio... Georgetown. Louisville, and Frankibrt, Ky. Chillioothe, Ohio.. Charleston, S.C... New York Detroit, Mich • « M • • UV ••••••■•••■« • • • • • Uw •••••• ••••■< • * • * • ^K\/ •••••••••••< • ■ ■ • • \aV ••*••••••••< do Pittsburgh, Pa.... Livingston Coan- ty.Ky. Bardstown, Ky... Geoigia • • * • • \AW •••••••■•••■ • ■ • • • Uw • «••••••*■*< Clarksville, Tenn. . Red River Washita River.... Middlebary,Vt... St. Genevieve. Mo. U. S. SaUne,in... Cmcinnati Date. Deo. 16, 1811 do do Dee. 17,1811 Dec. 16,1811 .do. 1 Dec. Dec. Dec. Dec. 17,1811 18,1811 to 30,1811 16,1811 Dee. 16,1811 Dec. 16,1811 do do — .do.... — — .do.. do do ...do... ,...do... ...do... ,...do... do... ...do... ...do... ...do.. Dec. 17,1811 Dec. 16,1811 do do Deo. 1&-Jan. 1 Jan. 1,1812 Dec. 16,1811 do do do do Jan. 23,1812 do... . ... do Jan. 23,1812 do Feb. 3.1812 Feh. 7,1812 do do do Feb. 8,1812 Feb. 7,1812 Feb. 8,1812 Dec. 23,1811 Feb. 7,1812 Feb. 16,1812 Dec. 14,1812 Dec. 16,1811 Hour. 8.15 a. m. . 2to3a.m 8 a. m.... Unoon.. 2.10 a. m. 2a.m. 2.30 a.m Early moming. 2to3a.m 2 a. m 2a. m 2.30 a. m Sunrise 2.15 a.m 2.47 a. m 3.34 a.m Daylight.... 8a. m 11.30a. m... 2 a. m 2 and 8 a. m. 2a.m 2 a. m. 3 a. m Daylight. 3.30 a. m (7). 2a.m 9 a. m... 9.20 a. m. 9.30 a. m. 9.15 a.m. 4.15 p. m. 4p. m... 7.30 p. m. 9.55 p. m. 11p. m.. 2a. m — 4 a. m — a. m a. m. a. m. a. m. a. m. 2.24 a. m. Remarks. Lig^t. Lasted 1 minute; motion enough to make Walking dlfflcalt; another soon af ter. Slight. Do. Cloeks stopped; objects feil; walls cracked; river agitated; trees waved. Banks of river caved; chimney thrown down. 48hocks. 1 or more daily. Lasted 3 minutes; Windows and f ur> niture shaken. One-half minute. Three slight shocks. Wakenedpeople. Movedfumiture. Severe shock. Violent; chimney wrecked. Do. Shook buUdings; chimneys thrown down. Do. As heavy as flrst. People lef t houses. Woke people. Overtumed nearly all chimneys. Losted 10 to 12 minutes; severe; buildings wrecked. Cradles rocked; bells rang; chimneys broken. Shocks daily. Bricks thrown from chimney. Shocks feit by Indians. Weak. Doubtful. Do. Several shocks feit. Similar to that of Dec. 16. 1 minute; buildings äiook. Books nearly thrown from shelve^ people stopped eating. FätT Do. Do. Severest feit. Reported by papers. Small. Lasted liminutes; nearly equaled that of Dec. 16. Small. Do. Do. Verücal motion. Alarming; severest yet feit. Horse reiused to proceed. Fissures and ejections. Severe shaking. Do. Do. Fairly severe. Principal shocks feit. Do. Shocks in winter. 500 shocks. Daily local shocks. Strong; lasted 6 to 7 minutes; moved fumituie; broke tops from chim- neys. 20 THE NEW MADRID EARTHQUAKE. Reoord of the süocka of the New Madrid eärthqiutke — Gontinued. Authority. Locality. Date. Hour. Remarks. Drake Cincinnati Deo. 16,1811 do 3 a. m SUsht vibrations. 7.20 a. m Modemt«^ rockinir, t^mi{natinF in .do do 7.30 a. m streng throe. Sligbt osdllations. ^0. ..... do do 10tolla.m 11.43 a. in .do Deo. 17,1811 Dec. 18,1811 Dec. 31,1811 3an. 3,1812 Jan. 23,1812 Jan. 27,1812 Feb. 4,1812 Feb. 5-«, 1812 Feb. 7,1812 Feb. 8,1812 do do ....do 11.30 a.m Moderate asitation. A few gentie rockings. 4to5a.in 2to3a.m 9 a, in do Slj^t vibrations. do . do ..•.••■■.... Oreat number of strong undulations in quick suooession, lasting 4 or 5 minutes. Solitary tremor as stronsr as that of 8.45 a. m ...do 4 p. m Jan. 23. A pretty strong agitation. Many sliffht jars determined by plumb Unes. Allday do 3.45 p. m do Allday last surpassing all previous shocks; chimney tops thrown down; fis- suies formed. Num^ronff verr sllirht tr^emors. do do 8 p. m Slight agitation. Vlbration oontinuing nearly a min- Ute. Strenger than last; much trembling, but little oscillation. Numerous slight tremors. Gentie Vibration. 8.30 p. m .....do do do ...do 10.40 p. m Nleht do Feb. 10,1812 Feb. 11,1812 do Feb. 13,1812 .... »QO» •*•«.. Feb. 16,1812 Feb. 17,1812 Feb. 20,18ir> Feb. 21,1812 Feb. 22,1812 Mar. 3,1812 Mar. 5,1812 Mar. 10,1812 Mar. 11,1812 Apr. 30,1812 May 4,1812 May 10,1812 June 25,1812 June 26,1812 Sept. 15,1812 Dec. 22,1812 Mar. 6,1813 Dec. 12,1813 do 41). m do do do ^ y 1 a. m Do. 6 a. m Do. 10 a. m Do. do do 2 p. m Do. 10a. ni(?) 3.40 a. m Do. do Strong shocks; undulation SSE. and Sli^t shocks. A short but strong shock. Slight south to north Vibration. A few slight rockings. Several short but strong rockings. Streng Vibration. Slight Vlbration. Moderate agitation. Slight shock. Do. do lOtollp. m 12.30 a. m do do 3to4a. m 6.30 a. in do do .. ..do 6.10 a. m 8 p. m .do 2to3a.ni (?) do . do y./..............> 11 a. m do 11 p. m do Nteht Slight agitation. 2 sfi^t vibrations. Do. do 8 a. m ..... do Daylürht do M^aj ta^Mm 3 p. m fllierht vihraHoTit do 10 p. m Very slight shock. Do. do 10 toll a.m 3 to4p. m do Do. The fact that shocks are shown for distant localities where none are indicated at New Madrid probably arises from the fact that no systematic or contmuous observations were made* at the latter point and only a few fragmentary Statements are available. It is, of course, a possibility that some of the more distant shocks were of local origm and unrelated to those of the New Madrid region. Other discrepancies arise from the fact that there were no seif- recording Instruments, and the lists indicate only those vibrations which could be feit or which happened to occur when the observers of the crude home-made pendulums were at hand. EFFECTS OÜTSIDE OF THE KEW MADBID ABEA. 21 BFFECTS OF THE SHOGKS OT7TSIDE OF THE NEW ICADEID ABEA. The following smnmaiy of the manifestations of the New Madrid earthquake, except those at Louisville and Cincinnati, is compiled maiiily from the excellent narrative of Mitchill.^ MISSISSIPPI VALLEY. The earthquake was feit at New Orleans, but was not severe. At Natchez four shocks were feit on the moming of December 16, the principal one being recorded as occurring at 2.10. Many houses were shaken, suspended objects swung to and fro, some plastered walls were cracked, a few articles feil from shelves, and several clocks were stopped. The surface of the river was agitated and parts of the banks feil in. The tops of the trees waved from side to side, but there seems to have been little or no noise. Near Piney River, Tenn., 20 acres of land adjacent to the river subsided until the tops of the trees were level with the surrounding earth. At Knoxville the first shock of the 16th is reported to have lasted more than 3 minutes, rattling the Windows and fumiture and awakening the inhabitants. Half an hour later another shock, lasting half a minute, was experienced, while between sunrise and breakfast three others, each of a few seconds' duration, were feit. At Columbia, in the same State, the people were awakened by the principal shock, which lasted 10 to 15 minutes. Itwas accompa- nied by a peculiar sound, which appeared to proceed from the south- west to the northeast. At Carthage and vicinity, also in Tennessee, there were one or more shocks daily from December 16, 1811, to January 1, 1812. At 3.30 on the latter date a shock threw bricks from chimneys and cracked the courthouse to its foundation. The motion appeared to proceed from south of west to north of east, or the reverse, and was greatest near the larger watercourses. In the shock of December 16 several chimneys were thrown down. At Clarksville, P. H. Cole reported the continuance of the earthquakes to December 15, 1812, one occurring on December 14. In Christian County, Ky., a fine spring became muddy and remained so for several hours. At the same time it became charged with hydrogen sulphide, presumably from the disturbance of decay- ing organic matter in the deposits from which it came. In Arkansas, according to Mr. Hempstead, a delegate to Congress from Missouri Territory, the roads between New Madrid, where court was held, and the Settlements of Arkansas, 200 miles distant, were rendered impassible by the earthquake. This made a circuit of 300 miles necessary, seriously interfering with the accessibility to the judiciary. 1 Mitchjll, S. L., A detaiied narrative of the earthquake which oocurred on the 16th day of December, 1811: Txaos. Lit. and Philos. Soc. New York, vol. 1, 1815, pp. 281-307. 22 THE NEW MADRID EABTHQUAKE. At St. Genevieve, Dr. Robertson, who then resided there, kept a record of the shocks until they amounted to more than 500, when he became weary of the task. At St. Louis, according to Mr. Riddick, the first shock was feit about 2.15 a. m. December 16, rousing people from sleep by the motion and the rattling of Windows, doors, and fumiture, to which was added a peculiar rumbling noise, resembling a number of car- riages passing over a pavement. At 2.47, 3.34, daybreak, 8, and 11.30 a. m. other shocks were feit, lasting from a few seconds to 2 minutes. Some chimneys were thrown down and a few stone houses split. OHIO VALLEY. At the time of the New Madrid earthquake Cincinnati and Louis- ville were among the largest Settlements west of the Allegheny Mountains and the only large towns near the earthquake center. As would be expected, therefore, it is from these places that the füllest details of the shocks are obtained. LOTTISVILLE. The record of the shock at Louisville was kept by Jared Brooks and was published by Henry McMurtrie in his sketches of Louis- ville.* Brooks constructed a number of pendulums of different lengths to detect the horizontal movements and a number of Springs to show the vertical vibrations. In this manner he was able to note and measure many vibrations not generally feit. His account in the Sketches gives many interesting details regarding the inten- sities of the shocks and the weather conditions. He also classified the shocks into six groups according to their intensities (see p. 33). The following table summarizes the shocks as recorded by Brooks in the publication mentioned. It wiU be f ound to difler somewhat in totals from the classified table of shocks, page 34, as, in the absence of the assignment of definite intensities in the text, it has not always been possible to determine the strength of a particular shock. Record of the earthquake shocks at Louisville.^ Dato. Dec. 1811. 16 17 18 19 Hour. 2.15 a. m. 2.30a. m., 7.20a. m. 5 a. m. 11.40 a. m. Strength. Violent. do Strong do. Strong to intense 6 oonsiderable shocks. Verysüght Weather. Cloudy and misty; temi>eratiire above freezing. Do. Do. Cloudy; some rain. Do. Gold; snowing. Winter weather. 1 McMurtrie, Henry, Sketches of lA)uJsville and its environs, Louisville, 1819. Appendix. 'CompÜed from notes of Jared Brooks. EFFECTS OÜTSIDE OF THE NEW MADRID AREA. 23 Record ofthe earthquake shocks at Louüville — Continued. Date. Hour. 1811. Dec. 20 21 ?2-28 29 30 31 1812. Jon. 1 2 3 4-8 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 1 2 Feb. 4 5 6 7 10.53 a. m.. 9 p. m 10.48a. m.. 8 to 12 a. m. 8to 12 p. m. 11 a. m. 4.05 a. m. 4.45 a. m. 12.21 a.m.. 9 a. m 9p. m 12.30 a. m.. 8 to 12 a. m. 3 a. m. 7 p. m , 7 a. m , 9 p. m , 9tol2 , lltol2a. m. 12noon 3 p. m , IIa. m , Night, a. m.. 11 to Inoon., 9 p. m.. 11 a. m. Night, a. m. Day 8.50 a. m Ip. m 10.30 p. m. . . Moming 11 to 12 a. m. 10.30p. m... Strength. Moming 8.50 a. m 9tol2a. m.... 9p. m 9tolla. m.... 11.30 a. m 9tol2a. m.... 3p. m 9tol2a. m.... P.m 9 a. m. tolp.m 2 a. m 9tol2a. m.... 12tol0p. m.. 10.45 p. m 10.45 to 11.30 p.m. 12.36 a. m 6 a. m 8tol2a. m.... ^. m A. m 3..30p. m 4.30 p. m 8.30 a. m 11.15 a. m 1.45 p. m 2.37 p. m 4.48 p. m Night, a. m . . . 8.30 p. m 3.15 a. m 3.15to5a. m.. 5 tol2a. m 12to8p. m... Considerable.. Slight tremors. ,....do do .do. More frequent tremors. Considerable Strong... Moderate. Considerable... Slight do Considerable... Strong tremors. Slight tremors.. Severe Slight tremors all day. Considerable Slight tremors do do Incessant tremors Considerable Slight ..r^do Tremors Incessant tremors Very f aint tremors Slight Considerable No shocks reported Very slight tremors. . , Slight tremors Considerable shocks. . . Very violent C. 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 24 25 26 27 28 3 8.10 p. m.. 10.10 p. m. 12.06 a.m. 12to6a 9.25 p. m 8.45 a.m 3.48 p. m 4.10 p. m 9.06 a.m 10.13 a.m 10.30 a.m lla.111. 11.50 a.m 3 p. m 8.25 p. m 5.40 a.m 6a. m 10a.m 12 m 0tol2a.iii.... 12to6a.m.... 9.19 a.m 10tolla.m.. 12 m 10tolla.m... 2.09 p. m 12.30 a.m 11.20 a.m 12.30 p. m 3p. m 9.15 to 12 a. m. 12m 10 p. m 4 a. m lltol2a. m... 3.07 p. m 2p. m 8tol0a. m.... Nig:ht, a. m... 10 p. m. 11.50 p. m Ni^t, a. m... 11.30 to 12 a. m. 8p. m 9a. m. P.m 12 p. m 8 to 12 a. m . . . 12m 4p.m 8.30p. m 12 p.m 9a. m Smirlae 9 a. m P.m 10.30 a.m.. P.m 8 to 10 a. m Smiset 4a. m 10 to 10.30 a.m. Ip. m Night, a.m... 8a. m 9.30 a. m 9.30 a. m. to 130 p. m. 6to7p.m 9 a. m 2p.m 2.35 to 3.30 p. m. 8.35 p. m SimrJse 8.35 a. m 2p.m 3J0p. m Bewtn to tRnModoos. Modenteahot^. Bmart shoA . . « . .....do. Do. Ooudy. Chrndy^or Fnaacnt aUglit ahodcB. . . «iSi!^:::::::::::; do Pendoloms all in motion. Laifely Tertical motioa. . Smart alMK^ Single bard shove Fabtahock Sljgfat motkMia .TTjio Gonaideiable ahock Sevcral tranofa ....do Cooaiderable tremor. Several tremof^ aljght... FUntabock SrreiBl faint tremora AH penduloma moved. . . . Gmaidcfable tremor SUght tremor. .TZdo Claarand Do. Do. Do. Do. Tbonder abowcr. Gafan and rainy. ParUydoody. Do. Do. Do. Do. do Continnal tremor. Conaiderabte motion Conalderable abock. Sbock of aome atrength Continnal tremor SUght abock Motion oooaiderable (over 70 ora dnring day and nisht). Pendnlnma awing half tne time . Tremors; few abocka dnring day Moderate abock do Do. Cloodj. Do. Do. Do. Rainy. Oeanng. Fair. Ckmdy. Do. Partly doady. trem- Frequent alisht vibrations. Pendnlnma m motion do. Peroeptible motion. . Freqnent dfebt moyements. Somewbat auurming aiMK^. Freqnent tremiMS , Considerable tremor. Smart Vibration. Strong -Vibration . do. 4-inch pendolum In motion (little motion throui^ day). Considerable trnnor Some motion Pendnlnma generally at rest All pendulnms in motion Little motion alter sunaet IrrM;u]ar movements of pendulnms . Slignt sbock (no motion during after- noon). Moderate sbock SUght to considerable motion Considerable sbock Smart motion Sensible shock Considerable sbock; considerable vertical motion. Tremors Pendulums in motion Pendolums move slightly SUght motion; stillest day yet as re- gardssbocks. Pendulums in motion; notioeable vibrations. Pendulums vibrate strongly Smart shock Strong shove and mudli motion SUght shock Vary sUght shook Pkaaant. Do. Do. Fair. Fineweather. CkNidy. Do. Do. Do. Do. Do. Do. Rainv. Do. Do. Do. Do. Do. Ck>ndy. Fair. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Do. Espedally fine weatber. Do. Do. Do. Pleasant. Do. Do. Do« EFFECTS OUTSIDE OF THE NEW MADRID AREA. 25 Record ofthe earthqujake shocka at Louisville — Continued. Date. Hoor. Mar. 1812. 5 6 8 9 10 11 12 Apr. 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 1 Simrise 10.301 5.25 p >. m. . m.. Evenlng Sunrise 10.30 12 m 3 p a. m. m. 3 a. m 11 a. m. to 2 p. in. 3 p. m 7.35 p. m 8 a. m 11 a. m 8.30 to 9 a. m. 9.36 a. m Noon 8p. m 9 p. ra Daybrcak 7 to 12 a. m . . 12to5p. m. Moming.... 6 to 9 a. m . 10.25 a. m.. Noon 3.20p. m... 7.20 p. m. . . Night, a. m. 9.20 a.m 10 a. m 12.25 p. m.. Night, a.m... 6.50 a. m 7.08 a. m 8.20 to 12 a. m. 6.30 p. m 8p.m Forenoon 2to3p.m 10.40 a. m. to 1 p. m. 9p. m 9 to 10 a. m 9 a. m 8 to 10 a. m Evenlng... 8.10 to 12 a. m. Aftemoon Sunset 8 a. m 9.20 to 12 a. m. 11.35 a. m. to 6 p. m. 12 p. m 8.25 a. m 10.05tolla.m. 3p. m 11.25 a. m 2p.m 12 p. m 7.20 a. m 7 to 8 a. m A.m P.m A.m , 2.20 to 7.10 p.m 7.30 a. m. Straogth. Shock generaUy feit Sadden loxig easy shove ConsiderablB moUon Sltght motions Shock generally feit A oonsfderable shove Very faint motion Conslderable horizontal and i>erpen- dicular motion. Oentle shove Barely peroeptible Vibration ....do Shock generally feit Slight Dut frequent tremors Pendulmns vibrate oonsiderably; conslderable perpaidJcular motion. Short pendulums move slightly Shock generaUy feit Perceptible shock Faint tremor Pendulums in motion Perceptible shock Pendulums in motion at short inter- vals. ....do 3 moderate shoves Pendulums seldom still Pendulums in motion All pendulums in motion Several pendulums in motion Powerfcu shoves in one direction 2 or 3 movements Strong shoves 3 to 6 inch pendulum^ in motion Short pendulumvibrates; littlemo- tion In aftemoon. Generally feit in town and oountry . . Short pendulums in motion All pendulums in motion Short pendulums in motion most of time. Barely perceptible vertical motion . . do No vibrations Conslderable shock; alarmed peopie. Medium-length pendulums in mo- tion. Slight motion Motion similar to preoeding day 3-inch vibrator in motion do Vibrators and pendulums in motion. The same during day and evening at intervals. Pendulums in motion Frequent slight motions 2 slight shocks 6-incn pendulum in motion Pendulum swings oonstantly Shorter vibrations; nearly oonstant. Strong shock 6-inch i>endulum moved oonsider- ably. All pendulums in motion ....do 2 and 4 inch i>endulums vibrate Similar to preoeding Considerable shock Similar to previous moming Less than previous day Still decreasing No vibrations do Pendulums vibrate Only Short pendulums move Peroeptible shock Pendulum vibrations Least motion of any day Several shocks suificient to move 4-Inch pendulum; 1 in night feit by peopie. Weather. Pleasant. Overcast. Cloudy. Rainy. Do. Dark. cloudy, oppressive. Clouay. Partly cloudy. Do. Do. Do. Do. Cloudy. Do. Fair. Cloudy. Partly cloudy. Do. Po. Hazy. Pleasant. Do. Fair. Do. Do. Do. Do. Do. Overcast. Cloudy. Do. Do. Approaching fair. Do. Do. Do. Do. Do. Do. Do. Fair. Do. Rain and thunder. Ralny. Clouay. Fine weather. Do. Fair. Do. Overcast. Hacy sun. Do. Ralny. Do. Snowing. Snowlng oocaslonally . Do. Calm, cloudy. Hasysun. Overcast, foggy. Do. Hasysun. Do. Ralny. Rain and lightnlng. Rain or snow. Fair. Fair and firosty. Do. Clear and crisp. Hazy sun. 26 THE NEW MADRID EABTHQtJAKE. Reoord of the earthquake sJiocks at Louisville — Gontmued. Date. 1812. Apr. May 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23-27 28 29 30 1 2 3 5 Hoiir. 11 a. m. Sunrise to 2 p. m. 6 a. m. 3p. m. P.m.. A.m.. 11 a. m. 8 a. m.. 9 a. m. Strangth. 6 to 10 a. m 10 p. m A. m 7 to 10 a. m P.ra 8 a. m. 8 a. m 10.25 a. m. 7.10a. in.. Similar to Apr. 1 Sensible shock Very slight tremors , Vibrations barely perceptible by pendolums. Frequent vlbratioxis of shorter pen- dmams. Motion at sunrise with slight vibra- tlons all dav. Perceptible snock Moderate shock Pendulums move at sunset Very slight movements of pendu- lums in moming. Perceptible shock Pendulums in motion Very slight movements do do ....do do Perceptible shock at daylight Pendulums in firequent action do Slight motion ....do Ck>nsiderable motion ....do Movements each day; generally slight, but some strenger. Perceptible shock before daybreak . . Streng vlbrations Strong vibrations in evening Motion moderate in moming, weak at night. Pendulums vibrate moming and evening. Many i>erceptible shocks; pendu- lums constantly in motion. Pendulums in motion Pendulums in motion moming and evening. Pendulums in motion Moderate to severe Frequent vibrations Weather. Dimsun. Fair. Fair to overcast. Cloudy and gloomy. Looked like gathering storm. Overcast, unsettled. Overcast, very dark. Overcast. Do. Partly cloudy. Cloudy, somo raln. Overcast, wind^. Overcast, but httle wind. Overcast. Do. Do. Do. Rainy. Heavy thunder storms. Rain and thunder. Showers. Do. Fair. Do. Fair and dry but with some haze. Calm, warm. Do. Do. Dry and calm; thunder shower in evening. Showers and thunder. Rainy. Do. Do. Threatening weather. Do. Fine weather. CINCnrNATI. A detailed account of the earthquake of 1811-12 as observed at Cincinnati is presented by Drake, who in f act left the only scientific record of the shocks at this point which the writer has seen.^ The chronological list of the disturbances was given in the table on pages 19-20. Of the first shock he says: At 24 minutes past 2 o'clock a. m. mean time the first shock occiirred. The motion was a quick oscillation or rocking, by most persons believed to be west and east; by some south and north. Its continuance, taking the average of all the observations I could collect, was 6 or 7 minutes. Several persons assert that it was preceded by a rumbling or rushing noise; but this is denied by others, who were awake at the com- mencement. It was so violent as to agitate the loose fumiture of our rooms, open par- tition doors that were fastened with falling latches, and throw off the tops of a few chimneys in the vicinity of the town. It seems to have been strenger in the Valley of the Ohio than in the adjoining uplands. Many feunilies living on the elevated ridges of Kentucky, not more than 20 miles from the river, slept during the shock; wbich can not be said, perhaps, of an} family in town. 1 Drake, Daniel, Natural and Statistical view or pictures of Cincinnati, Cincinnati, 1815, pp. 23ä->244. EFFECTS OtJTSIBE OF THE NEW MABBID AREA. 27 Qf the severe shock of Januaiy 23 he says : About 9 o'clock a. m. a great number of strong undulationB occurred in quick suc- cefision. They continued 4 or 5 minuteB, having two or three distinct exaacerbatioiiB during that tdme. An Instrument constructed on the principle of that used in Naplee, at the time of the memorable Galabrian earthquakes, marked the direction of the undulations from south-southeast to north-northwest. This earthquake was nearly equal to that which commenced the series on the 16th ultimo. Hüs account of the severe shock of February 7 is as foUows: At 45 minutes past 3 o'clock a. m. several alarming shocks in rapid succession. The Instrument already mentioned indicated the three principal heaves to be from the Southwest, the south-southwest, and southnaoutheast. The last greatly surpassed any other undulation ever known at this place. It threw down the tops of more chimnejrs, made wider fissures in the brick walls, and produced vertigo and nausea in a greater number of people, than the earthquakes of either the 16th of December or the 23d of January. It was said by some that this earthquake was preceded by a light and a noise; but this was denied by others, who were awake and coUected in mind and senses. On February 8 the following occurred: At 40 minutes past 10 o'clock a shock considerably strenger than either of the pre- ceding [of February 8]. It was observed to produce in suspended and elevated bodies a very sensible degree of trembling, but no oscillation; indicating, perhaps, a vertical instead of the horizontal motion of the previous shocks. Immediately before this shock I had the satisfaction of hearing, for the first time, a noise such as preceded, according to the report of some of our Citizens, most of the principal earthquakes. It was a peculiar, faint, dull, rumbling or rushing sound, near the horizon, to the Southwest. It seemed to approach but not arrive at the place of Observation, and after continuing four or five seconds was succeeded by the shake. During the remainder of the night and the next day the earth was in the same State of tremor which it suffered on the 5th and 6th. Discussmg the mtensities of the shocks, Drake adds: The violence of diSerent earthquakes is best indicated by their efficiency in alter- ing the structure of the more superficial parts of the earth, and in agitating, subvert- ing, or destroying the bodies which they support. On a comparative scale, formed from such remarks, at this place the first shock of the 16th of December, 1811, that of the 23d of January, 1812, and the first on the 7th of February, occupy, above the rest, a decided elevation, and constitute the first class. To the second class belong the ehock at 20 minutes past 7 o'clock a. m., December 16, that on the 27 th of January, and that at 40 minutes past 10 o'clock p . m. on the 8th of February. Of the remainder, one-half, by estimation, may be referred to a fourth class, composed of those which were feit only by persons not in action; and the remainder will constitute a third class of intermediate violence. The numerous tremors and ebuUitions that were detected by pendulums, and the delicate sensations of a few nice observers, when at perfect zeet, may constitute the fifth and lowest order of these multiplied agitations. After giving the description quoted above, Drake presents a de- tailed account of the physical phenomena observed during this period of disturbance and during previous earthquakes. (See p. 13.) OTHER LOCALITIES. At Jeflfersonville, opposite Louisville, the shocks were feit very inuch as they were in Louisville. At Red Bank, 150 mUes below 28 THB KEW MADBm IIABTHQTTAKE. Louisville, Anthony New stated that from December 16, 1811, to January 4, 1812, there were 20 to 30 shocks, the one shortly after 2 a. m. and the one at sunrise on the 16th being the most violent. These shocks, which cracked or threw down many chimneys and caused the people to flee from their houses, are said to have been accompanied by a noise coming from the west. At Vincennes, on the Wabash River in Indiana, the first shock was very severe, menacing the safety of the honses, and was followed by shocks of less vioienee for several days. At Lebanon, Ohio, the vibrations, which appeared to move from east to west, were so great that the houses were vacated. At Circleville, in the same State, a violent trembling sufficient to bring persons from their beds was experienced at the time of the first shock, while a lesser shock was feit at 8 a. m. In the Green River region and Henderson County, Ky., the first shock, which was alarming in its nature, was followed through night and day up to December 30 by less violent shocks. On the latter date a shock more severe than any of those preceding was experienced, over- tuming nearly every chimney in Henderson Coimty. The shock of 4 a. m., February 7, was severe at Pittsbm'gh, being greater than any previously experienced. Many persons left their houses. GREAT LAKES REGION. Judge James Witherall, writing from Detroit, Mich., reported that the first shocks were distinctly feit in that region. The weather was cold but calm. In February, 1812, he wrote that the earthquakes still continued. On the 3d instant, 14 minutes past 4 p. m., a small shock was feit; the mercury low, but not quite in the ball; it had risen very considerably a few hours previons to the shock. On the 7th, at 4 p. m., the weather continuing moderate, the shock was strong, nearly equaling the one I previously mentioned to you, and continued about 90 sec- onds; on the same evening at half-past 7 another small shock; at 55 minutes past 9 the same evening another small shock; at 11, another; and at 12 a. m. of the Sth, one which seemed to produce a different motion, that is, like pounding up and down instead of oscillating. A. B. Woodward, another judge in Michigan Territory, wrote on April 7: We have had nine shocks of the earthquake here, of which I have an exact memo- randum of eight and have somehow entirely lost the time of the other. I feit four myself. I know only one person, a French lady, who feit the whole; speaking here of the eight. In a letter dated June 23, the same gentleman observes: In a late joumey to the Rivi^re auxTranches, in upper Oanada, I found the number of shocks of the earthquake feit there was exactly the same as here — that is, nine. EFFECTS OUTSIDB OF THE NEW MADRID AREA. 29 ATLANTIC COASTAL AND PIEDMONT PLAINS. The shocks were distinctly feit throughout the Piedmont and Coastal plains from Washington southward, but were not very notice- able farther north, although reported at Baltimore and even at Boston. The nontransmission of the shocks in the Northern States was probably due in part to their distance from the centrum and in part to the oblique angle made by the line of transmission of the vibrations to the strike of the rock strata, the shock apparently being transmitted much more readily in directions at right angles to the trend of the strata, as in the southem Appalachians, than in those making very low angles, as at the northem end of the System. The following include the principal allusions found in regard to the shocks.^ WaskingUm, — On the moming of Monday, the 16 December, 1811, several shocks of earthquakes were feit at the city of Washington. The first of these happened at 3 o'clock, and in some houses was considerable, enough to shake the doois and Windows and wake persons from their sleep. There were successive tremore. Tasseis of cur- tains were seen to move, and pitchers of washing Stands were heard to rattle upon their basins. The sound was very distinguishable and was believed by many to pass from Bouthweet to northeast. The alarm was so great in some families that searches were made from room to room to discover the robbers who were imagined to have broken into the houses. A second shock, though lighter, was experienced about 6 o'clock, and a third about 8. A gentleman standing in his Chamber at his desk and writing, in the third story of a brick house upon the Capitol Hill, suddenly perceived his body to be in motion, vibrat- ing backward and forward and producing dizziness. Not suspecting at the moment that the uncomfortable Sensation was caused by an earthquake, he examined his desk to know whether it stood firm. Finding that it did, he dropped his pen and, tuming his eyes upward, discemed that the looking-glass and other things hanging near him were in a similar motion. Another person was near a table placed beneath a mirror. Feeling a giddiness come u}>on bim, he seized the table for support. The general agitation of the Chamber and house ceased in about a minute; but the looking-glass, which was suspended in the usual manner, continued to swing for some seconds longer. Points in the South. — In Richmond the signs of an earthquake were witnessed by many persons. At 3 o'clock on the same moming (the 16th of December) there were Said to be three successive shocks, another about 6, and a third about 8. Several people were impressed with a belief that thieves had entered their dwellings, and in one of the most elevated mansions the bells were set a-ringing in both the upper and lower rooms. The noise and concussion were supposed by some to proceed from east to west. It was stated at Norfolk that two very distinct shocks were feit in that town and in Portsmouth, to wit, at 3 and 8 o'clock in the moming of the 16th. Some clocks were reported to have stopped, the doors rattled, and articles hanging from the ceilingB of shops and houses swung to and fro, although a perfect calm prevailed. At Raleigh, N. C, several slight earthquakes were feit on the moming of the 16th of December. The first happened between 2 and 3 o*clock and was distinctly perceived by all who were awake at the time. Two others were reported to have occiured between that time and 7 o'clock, but were not plainly observed except by some mem- bers of the legislature who were in the Statehouse and were considerably alarmed at the shaking of the building. 1 Mltohill, S. L., Trans. Llt. and Philos. Soc. New York, vol. 1, 1815, pp. 281-307. 80 THE NEW MADBID EABTHQUAKE. From Geoigetown, S. C, it was told tliat several ahocks had been experienced between the houis of 3 and 8 on the moming of the 16th. The inhabitants weie much alanned. The shocks were so considerable that the parade gioimd of the fort was said to have settled ffom 1 to 2 inches below its fonner level. A tub of water standing upon a table in the barracks was reported to have been overset by the jarringof thebuilding. Another severe shock was feit two days afterwards at noon. At Columbia, S. C, the inhabitants were alarmed by repeated fihocks. The fiiBt took place at half after 2 in the moming of Monday, which was represented as shaldng the houses as if rocked by the waves of the sea. It was followed after the cessation of a minute by three slighter ones. At 8 o'clock two otheis took place, and at 10 some slight ones. The South Carolina College appeared to rock from its foundation and a part of its plaster feil, which so alarmed the students that they left their Chambers without their clothes. It seemed as if all the buildings would be leveled. The dogs barked, fowls made a racket, and many peisons ran about with lights, not knowing where to go so great was their terror. During the first agitation it was observed that the air feit as if impregnated with a vapor, which lasted for some time. At Charleston , S . C . , the Sensation was of considerable strength . One account stated that on the moming of the 16th, at a few minutes before 3 o'clock, a severe shock of an earthquake was feit. Its duration conjectured to have been between 2 and 3 minutes. For an hour previoiis, though the air was perfectly calm and several stans visible, there was at intervals of about 5 minutes a rumbling noise like that of distant thunder, which increased in violence of sound just before the shock was feit. The Vibration of St. Fhilip's steeple caused the clock bell to ring about 10 seconds. Two other shocks were feit afterwards, one a little before 8, and the other about a quarter of an hour after. Both these were slighter and shorter than the first. Many of the family clocks were stopped by the concussions. In many wells the water was considerably agitated. From another source it was related that Charleston was shaken by an earthquake severely at the time before specified. This was preceded by a noise resembling the blowing of a smith's bellows. The agitation of the earth was such that the bells in the church steeples rang to a degree indicative of an alarm of fire. The houses were so much moved that many persons were induced to rise from their beds. The clocks generally stopped. Another slight shock was experienced about 15 minutes after, and yet another at 8 o'clock. This last one produced a considerable rattling among glass, China, and other fumiture. A looking-glass hanging against a west wall was observed to vibrate 2 or 3 inches from north to south. The inhabitants of Savannah were sensible of four earthquakes. The fiist was on the moming of the 16th of December between 2 and 3 o'clock. It was preceded by a flash of light and a rattling noise resembling that of a carriage passing over a paved road. It lasted about a minute. A second soon succeeded, but its duration was shorter. A third happened about 8 o'clock, and a fourth about noon on the 17th. Per- sons who experienced the hardest shock were made to totter as if on shipboard. Its course was believed to be from southwest to northeast. It was observed by Dr. Macbride, of Pineville, S. C, that the earthquake terrified the inhabitants exceedingly. It was accompanied by several appearances that coun- tenance the theory of this phenomenon which brings in the agency of the electric fluid. (1) The infrequency or absence of thunder storms; that is, they were much less frequent this year than usual, especially in the autumn. (2) Immediately before the earthquake a red appearance of the clouds which had much darkened the water for 24 houiB immediately before the shock. (3) The loudness of the thunder, and the number of the peals within 24 hours after the first shock and but a few hours before the last, which was feit before he wrote. Such thunder was very imusual at that season. The Hon. Israel Pickens, of Buncombe Coimty, N. C, received a letter from the Rev. John Carrigan, dated the 28th of February, and containing the foUowing informa- tion: NATURE OF THE VIBRATIONS. 31 ''Durin^ my travels lately to and from the State of Geoigia I made it a part of my busmess to obtain the most accurate accounts of the present shakiiig of the earth from all parte. I have found that in all parts of the Continent the motion of the earth has been the same and its partiality remarked in the same neighborhoods. In this country the first rocking (as it is generally called) was perceived on the 23d of December^ a little bef ore daylight. Since that time it has been observed almoet every week throngh South Carolina and parts of Geoigia. Several persons in those States have told me that they have feit it almost every day since. No damages have arisen more than a few bricks shaken off some chimneys. There is no truth in the report in circulation respecting the fall of the Painted Rock and other extraordinaries in Buncombe Connty. I gave my friend, Ool. Freeman, in Georgia, a call, who informed me that he had par- ticularly noticed some tall poplars in his lane during the time of the second shock rock- ing with an equable motion from northeast to southwest, which I have found to be general. On the 7th and 16th instant the shaking has been general here.'' NATTJBE OF THE VIBRATIONS. In the New Madrid area proper the only evidence of the nature of the vibrations is the accounts of those persons who experienced them. Of these Bringier, who experienced one of the major shocks, describes the action as a blowing up of the earth accompanied by loud explosions.^ Casseday writes: * ''It seems as if the surface of the earth was afloat and set in motion by a slight appUcation of immense power, but when this regularity is broken by a sudden cross shove, all Order is destroyed, and a boiling action is produced, during the continuance of which the degree of violence is greatest, and the scene most dreadful." Others described the movement as an ''undulationof the earth re- sembling waves, increasing in elevation as they advanced, and when they had attained a certain fearful height the earth would burst/' * Le Sieur likewise says: ''The earth was observed to roll in waves a few feet high with visible depressions between. By and by these swells burst throwing up large volumes of water, sand, and coal.'' * Hajrwood writes that in Tennessee the motions were undulating, the agitated surfaces "quivering Uke the flesh of a beef just killed." The motion is said to have "progressed from west to east and was sometimes, though seldom, perpendicular, resembling a house raised and suddenly let fall to the ground.'' ^ Audubon, who experienced one of the shocks while riding in Ken- tucky, says that "the ground rose and feil in successive furrows like the ruffled waters of a lake. * * * »pj^^ earth waved like a field of com before the breeze." ^ There were great differences in the intensity and destructiveness of the shocks. According to Fester: ^ "Sometimes they would come » Bringier, L., Am. Jour. Sei., Ist ser., vol. 3, 1821, pp. 15-46. s Casseday. Ben., History of Louisville, 1852, p. 122. s Flint, Timothy, Recollections of tbe last ten years, Boston, 1826, p. 223. * Le Sieur, Oodfrey, quoted in Am. Geologist, vol. 30, 1902, p. 80. * Haywood, John, Natural and aboriginal history of Tennessce, NashviUe, 1823, p. 124. * Audubon, J. J., Audubon and his Journals, New York, 18d7, yol. 2. pp. 234-237. » Fostor, J. W., The Mississippi Valley, Chicago, 1869, p. 20. 82 THE NEW MADBID BARTHQUAKE. on gradually and finally culminate; again they would come without premonition and in terrific force and gradually subside." When the shocks were severe it was practically impossible to stand. Bradbury and bis crew, who encountered one of the severe shocks while on shore caring for bis boat, had great diflSculty in preventing them- selves from being thrown down, and one of the party was nearly precipitated into the river.* Again Flint says that when the shocks ''were at the severest point of their motion the people were thrown on the ground at almost every step. A French gentleman told me that in escaping from bis house, the largest in the village, he found he had left an infant behind, and he attempted to mount up the raised piazza to recover the child and was thrown down a dozen times in succession." ^ Vertical shocks do not appear to have been lacking in the New Madrid area, but, according to Flint, were less destructive than the horizontal type. He says: ''The shocks were clearly distinguishable into two classes, those in which the motion was horizontal and those in which it was perpendicular. The latter were attended with the explosions and the terrible mixture of noises, * * * but they were by no means as destructive as the other." ' In the more remote districts the action was less intense, producing only vibrations and tremors. There appears, however, to have been more or less of surface movements, as the shocks were much more distinctly feit by those living in the alluvial flats of the Valleys than by those on the rock uplands, notwithstanding that it is only through the rocks that the shocks could be transmitted to the distances observed. The slight vibrations in the latter must, therefore, have been greatly magnified on transmission to the alluvial masses. The intensities in Valley and upland differed sufficiently to be noticeable at the time. Drake, speaking of Cincinnati, says:* "The convulsion was greater along the Mississippi, as well as along the Ohio, than in the uplands. The strata in both Valleys are loose. The more tenacious layers of clay and loam spread over the adjoining hills, many of which are composed of horizontal limestone, suffered but little derangement." At Louisville Jared Brooks constructed a number of pendulimas varying from 1 to 6 inches in length. According to his records, as presented in McMurtrie's history,* although there was great Variation in the rapidity and amplitude of the vibrations at different times, there was often considerable uniformity during a single shock. Sometimes when the amplitude was short only the 1 or 2 inch pendu- lums would vibrate. At other times the 3 and 4 inch pendulums 1 Bradbnry, John, Early westem travels, Cleveland, 1904, vol. 5, p. 257. s Flint, Timothy, op. cit., p. 225. * Op. Ott., p. 224. * Drake, Daniel, Natural and Statistical view or picture of Cincinnati, Cincinnati, 1815, p. 235. * MQÜwrtiiß, H., Sketches of Louisville and its environs, Louisville, 1819, Appendix. CLASSIFICATION OF INTENSITEES. 33 would move while both the longer and shorter ones remained sta- tionary, and again the 6-inch pendulum alone would be in motion. The greät majority of the vibrations at Louisville appear to have been horizontal and to have varied from mere tremors to streng movements to and fro. Usually a considerable number of move- ments oceurred dnring each shock, but at times the motion was like a shove in a single direction without any corresponding move- ment in the opposite direction. Sometimes two or three such shoves oceurred in succession, varying in intensity from gentle to sharp. A few motions were vertical and were recorded only by the spring indicators, the pendulums remaining quiet. Other notes on the movements are given on pages 22 to 26. COliASSIFICATION OF INTENSITIES. No record of intensities was kept at New Madrid or vicinity, but at Louisville, only a short distance away, a systematic record was kept by Jared Brooks, who divided the shocks into six classes, viz: Fiistrate. Most tremendous, so as to threaten the destniction of the town, and which would soon effect it should the action contmue with the same degree of violence; huildings oscillate largely and irr^^arly and grind against each other; the walls split and begin to yield; chimneys, parapets, and gable ends break in various direc- tions and topple to the ground. Second rate. Less violent but very severe. Third rate. Moderate but alanning to people generally. Fourth rate. Perceptible to the feeling of those who are still and not subject to other motion or sort of jarring that may resemble this. [Fifth rate. Not defined.] Sixth rate. Although often causing a stränge sort of Sensation, absence, and some- times giddinessy the motion is not to be ascertained positively, but by the vibratoiB or other objects placed for that purpose. In his classified list he records a total of 1,874 shocks between December 16, 1811, and March 15, 1812, but many more oceurred in the succeeding months for over a year, as shoMm in the tables on pages 25-26. Of the 1,874 shocks, 8 are classed as violent, 10 as very severe, 35 moderate, 65 generally perceptible, 89 of ''fifth rate," and 1,667 indistinctly feit or noted by movements of delicately poised objects. (See table on p. 34.) PEBIODS OF ACTIVITY. The greatest intensity of action oceurred not at the beginning of the disturbances but in the first half of the foUowing February and continued through the later part of February and the early part of March. The activity beginning on December 16 gradually increased imtil the middle of January, but feil off during the later part of the month imtil only a little over one-third of the former number of shocks was feit. Then followed a second period of great 34 THE NEW u fAnttm KABTHQVAKB. activity in late Januaiy and early Febniuy and aftor a falling ofi in the middle of the month a period of still greater activity late in February. After another falling off in early March a further period ol intensity developed sbout the middle of the month. This is brought out by the lioes in £gure 2. Figure 3 e^ows similar lines for shocks of ihß first, second, third, and fourth intensities. It is interesting to note that the linea for the heavy shocka, or those of intensities from 1 to 3, are Bynchronous, but that of the fourth intensity tends to be highest when the otfaers are relatively low. Thia sug- gests that instead of readjustr ments by a few lai^ shps (intensities 1, 2, and 3) the change is at times by a tai^ niimber of smaJl slips, giving rise to very faint shocks, fl ^ @ such as those of intensity K 4. Shocks in the weeks end- ing December 22 and Feb- ruary 9 fumish examples of FioTTBB 2.— Disgram showtng evthimake activit? by readjustment by a few lai^e woets.bftae THE WEATHEB. 39 Distrümtion ofahocka wüh reference to phases ofthe moon, [Based on table of Jared Brooks, npiinted on page 34.] Week ending. Dec.22. Dec.29. 1811. Jan. 5.. Jan. 12. Jan. 19. Jan. 26. Feb.2.. Feb.9.. Feb. 16. Feb. 23. Mar.l.. Mar. 8.. Mar. 15. 1812. Phase of moon. Phase. New. FuU. New. Füll. New. Fall. New. Date. Dec. 15 Dec. 29 Jan. 13 Jan. 27 Feb* '12' Feb. 26 Mär."i2' Oeneral phase for week. First -quarter... Seoond quarter. Third quarter. . Fourth quarter. First quarter. . . Second quarter. Third quarter. . Fourth quarter. First quarter. . . Second quarter. JThird quarter. Fourth quarter. . . . Total shocks. { 87 156 134 161 65 91 209 175 86 292 139 58 221 Snnunary for qnarten. First quarter. Second quarter. Third quarter. Fourth quarter. DflcembeT-Jannary 87 65 86 156 91 292 134 209 197 161 Jrninary-^f^bnwry 176 FpbniMy-March T 221 238 539 540 557 RELATION OF THE EABTHQXJAXE AND THE WEATHEB. There has been for some time a growing tendency to regard specula- tions as to the relations of earthquakes and weather as futile. So far as this applies to temperature and precipitation the tendency is probably warranted, but there appears to be reason to believe that the variations of pressure associated with cyclonic or storm moye- ments are of more importance than is sometimes imagined. Varia- tions of an inch in the length of the mercury column, which are by no means excessive, mean a change in pressure of half a pound per Square inch^ or 3^ per cent of the total atmospheric pressure. Such changes in pressure, especially when they are sudden, have profound effects. The water in deep wells often fluctuates materially with changes of barometric pressure; many wells flow duringthe passage of cyclonic Centers which at other times fail toreach the surf ace; broad tidelike waves (seiches) several feet in height are not infrequently formed in the Great Lakes; and recent observations by Omori* near Tokyo, Japan, show that the passage of cyclonic storms sometimes even pro- duces measurable tiltings of the earth's surf ace, the ground rising and falling as the cyclones progress. 1 Nature, Sept. 26, 1907, p. 553; Scienoe, new aer., voL 26, p. 761. 40 THE NEW MADBID EABTHQUAKE. No record of barometric pressures in the vicinity of New Madrid at the time of the earthquake seems to be available; but in view of the possible importance of the relation of earthquakes to variations of atmospheric pressure, it is pertinent to present the inferential evidence aflforded by the general weather conditions in adjacent regions. Very complete notes on the general weather conditions at Louisville were made by Jared Brooks.* These are abbreviated in the table on pages 22 to 26. From these it appears that the weather, although marked by long cloudy or stormy periods, was really not abnormal unless somewhat warmer than usual. In the time for which the record was kept, extending from December 16 to May 5, there were 100 fair, 42 cloudy, and 33 stormy days. A study of the shocks separately recorded shows that 89, or 0.89 per day, took place in pleasant weather, 100, or 2.38 per day, in cloudy weather, and 66, or 2 per day, during storms. The cloudy and the rainy weather should, however, be combined in a single class, since rain is a mere and uncertain incident to the cloudiness. This grouping gives 89 shocks, or 0.89 per day, in fair weather, as against 166, or 2.21 per day, in cloudy or stormy weather. The details of the shocks, as brought out more fully in the foUowing table, seem to indicate that about twice as maay shocks occurred in periods of low barometer (as indicated by cloudiness and rain) as in periods of high barometer, while the average number per day was nearly two and one-half times as great in stormy and unsettled weather as in fair weather. Relation of shocks to the weather, as ohserved at Louisville, Ky. [Based on table, i>ages 22 tö 26.a] Intenalty. Intense, vlolent Strong, severe Moderate, considerable Strong yibrations, considerable vlbrations, slight shocks Falnt shocks, tremors, weak vibrations.... Fair (100 days). Number of shocks. 2 5 13 23 46 89 Number of shocks daUy. 0.02 .06 .13 .23 .46 .89 Cloudy, etc. (42 days). Number of shocks. 2 1 16 21 60 100 Number of shocks daily. 0.05 .02 .38 .50 1.43 2.38 Stormy (33 days). Number of shocks. 4 4 15 13 30 66 Number of shocks daUy. 0.12 .12 .45 .40 .90 2.00 o The lack of agreement as to the total numbers of the less severe shocks between this and the summaiy table of Brooks (p. 34) is due to the indefiniteness of some of the descriptlons in Brooks's llst. DIBECTIOX OF VIBItATIOXS. The reports of the direction of vibrations are much less complete than could be wished, and of those available only a few rep'^esen.t instrumental determination, and even these were recorded on more or 1 McMurtile, H., Sketches of Louisville and its enviroDS, Louisville, 1819, Appendix. DIBECnON OP VIBRATIONS. 41 less imperfect home-made instruments. The recorded directions of the principal shocks are presented in the following table. The direc- tions are also shown graphically in figure 1. The apparent dh*ections of sounds are omitted as being too unreliable for record. The sounds appear to be produced by elastic waves of the same origin as the more powerf ul waves giving rise to the earth tremors and seem to be trans- mitted through the groiind and communicated to the air just before they reach the ear. The general directions of the air yibrations would thus be upward and downward, giving rise to very illusory impressions. It is interesting to note that vibrations capable of producing sound reached as far as Washington, D. C, and Charleston, S. C. Directions from which vibratUms of the New Madrid earthquake were reported as Coming. Locality. Authority. Dec. 16, 1811. Dec. 17, 1811. Jan. 23, 1812. Feb. 7. 1812. Feb. 8, 1812. Feb. 17, 1812. Feb. 22, 1812. New Madrid, Mo... Varioos... Mitohill... ...do ...do Brooks Drake Mitchlll... ...do . ..do...... . SW. S£. SW. SW. f SW.i,« S.i»« lVert.1»« W. E. "'S.'c'?)"' Heiculaneaxn, Mo.. Columbia, Tenn Carthage, Tenn LouisvUle, Ky * Clncinnatl, Ohio. . . SSE.i f SW.i SSW.i SSE.i Vert.i SSE.i S.i Lebsuioii, Obio Detroit, Mich SW. Vert.i Charleston, S. C Ravnnilllb, Oa. . ...do SW. ...do SW. ^ Instrumental determinations. >Most. «Few. 4 Oocasional. From the table it appears that the direction of the vibrations at New Madrid was from the west, while at Hercnlanenm, which is northwest of New Madrid, the movement was from the southeast. At Red Bank and Cincinnati, northeast of New Madrid, the vibra- tions were from the west; but at Lebanon, not far from Cincinnati, they are reported to have been from the east. At Carthage and Columbia, Tenn., nearly due east of the area of greatest disturbance, the direction of the vibrations was from the southwest. At Wash- ington, D. C, the movement was reported as northeastward. Considering the düficnlty in determining the direction from which a shock comes, even when the trend of the oscillation is known, it is perhaps surprising in the absence of any recording instruments that the records agree as well as they do. An example of this was presented in Lebanon, Ohio, where the movement was in an east- west direction, but whether propagated from the east or from the west could not be readily determined. It was reported as coming from the east, though it came, in all probability, from the west. In fact, except that from Lebanon, the observations on the direction of the first shock are all reasonably accordant, the slight Variation ehown at the Ohio Biver towns doubüess being due to local influence. 42 TECE NEW MADEID EABTHQUAKE. The direction of the shock of December 17 was recorded only at Detroit and Savannah; although it was feit at other places. The direction of the yibrations at the former city were normal f or a shock originating in the New Madrid area^ but at the latter city were some- what abnormal, possibly owing to a local tendency to move in the line of least resistance toward the adjacent river. The shock of January 23 was recorded by Instruments at Cinciii- nati as from the south-southeast, which must be regarded as abnormal if the shock originated iq the New Madrid area, but would be normal for a readjustment taking place in the Appalachians. It should be noted, however, that the shock was comparable in severity to that of December 16 in the New Madrid area, indicating that the focus could not have been very remote. The shocks of February 7, often known as the big shocks, were three in number and were recorded in Charleston as from the South- west, south-southwest, and south-southeast. The two former are normal for a shock originating in the New Madrid area. The shock of February 8 is notable for the apparently vertical character of its vibrations, which are reported from points as remote as Cincinnati and Detroit. No record of a shock on this date at New Madrid has been found, although very Hkely there was one. The same is true of the shocks of February 17 and 22. EFFECT OF THE EABTHQX7AKE ON LIFE. As would be supposed from the severity of the disturbance, human bemgs nearly everywhere were terrified by the earher shocks, their behavior giving rise to many ludicrous and many pathetic incidents. Even the lower animals shared the f eelings of great alarm. The most graphic accoimts of the f eelings and behavior of the inhabitants during the shock are given by Drake, Mitchill, Foster, Bryan, LeSieur, Audu- bon, Wetmore, Carr, Bradbury, Brooks, and Cassedy. The eflfects on lower animals are described by Audubon, Bryan, Mitchill, Bradbury, Hildreth, and Fhnt. (See Bibliography, pp. 111 to 115.) Later, bb the inhabitants became more familiär with the shocks, they lost much of the fear and began to observe the phenomena more closely. In Louisville and elsewhere swinging and spring pendulxuns were set up for the purpose of detecting the minor tremors and of distinguishing the horizontal and vertical movements. Jared Brooks, especially, conducted a long series of experiments, using pendulmns from 1 to 6 inches in length, the results of which were recorded fully. (For sum- marj', see pp. 22 to 26.) The New Madrid earthquake caused very sHght loss of life, as com- pared with most other earthquakes of similar intensity. Thus, at Caracas, which was destroyed during the continuance of the New Madrid shocks, over 10,000 were killed by falling walls^ whereas in GOVEBNMENT BEI/IEF. 43 the New Madrid axea only a single life has been reported as lost in this manner. This immimity was due to the f act that the region of greatest disturbance was thinly settled and without large towns with brick or stone buildings to coUapse and crush the inhabitants. In this conntry region the buildings were mainly cabins, which were strongly biult and at the same time more or less elastic. The frame houses of the larger villages were likewise of a character to give before rather than resist the vibrations. The fact that the people left their houses at the first shock and were in places of compara- tive saf ety when the later and more severe shocks came also has much to do with their escape. The only death on land of which accounts have been seen was that of one woman who, frightened by the shock, ran until her strength gave out and expired of fear and exhaustion.^ One writer teils of others who were thrown into the river by caving banks at New Madrid and drowned.^ None of the contempora- neous accounts, however, seem to mention the latter incident. An account of a loss of life by the caving and disappearance of Island No. 94, near Vicksburg, is given by Broadhead.^ Quoting from an old account of the trip of Capt. Sarpy, of St. Louis, he says: They tied up at this island on the evening of the 15 th of December, 1811. In looking around they found that a party of river pirates occupied part of the island and were expecting Sarpy with the intention of robbing him. As soon as Sarpy found that out he quietly dropped lower down the river. In the night the earth- quake came and next moming when the accompanying haziness disappeared the island could no longer be seen. It had been utterly destroyed as well as its pirate inhabitants. On the river a number were drowned. Bradbury mentions seeing drifting canoes, the owners of which he afterward found had been lost.* Hildreth describes the loss of several boats and their crews by caving banks.* Lloyd ^ records that a flat boat belonging to Richard Stump was swamped and six men drowned. Many other boats were destroyed by snags and the river was covered with wrecks. So numerous were the disasters that the escape and arrival at Natchez of Capt. Roosevelt with the steamer New Orleans was regarded as almost miraculous. QOVEBNMEKTT RELIEF. The earthquake, as has been indicated, left many persons home- less, while the habitations of others were rendered so dangerous that the owners were compelled to take up their abode in tents or temporary huts for safety. Many farms were in whole or in part 1 Flint, Timothy, Recollections of the last 10 years, Boston, 1826, pp. 222-228. »Lloyd, J. T., Lloyd's steaznboat directory, Cinctnnatl, 1856, p. 325. s Broadhead, O. C, The New Madrid earthquake: Am. Oeologist, vol. 30, 1902. « Bradbury, John, Early westem travels, Cleveland, 19U4, vol. 5, p. 206. » HUdieth, S. F., Original contributions to the American Pioneer, Cincinnati, 1844, pp. 34-35. 44 THE NEW MADBID EABTHQUAKE. precipitated into the streams; others were covered by extnided waters; still others were depressed and gradually covered by accu- mulating surface waters. The aggregate loss of tillable land reached thousands of acres. Moreover, the continuance of the shocks gave rise to an uneasiness for the future which seriously retarded the recovery. The loss and suffermg were eventually brought to the attention of Congress, but in the light of subsequent events it is not certain to what extent assistance was the real object of the agitation or to what extent it was a pretext for land grabbing on the part of certain unscrupulous persons. The agitatiou in Congress resulted in the passage of an act aUowing those whose lands had been destroyed by the earthquake to locate the same quantity of land in any other part of the pubhc domain open to entry. Few of the losers took advantage of this, however, for by this time the shocks had nearly ceased^ the people had become ad- justed to the new conditions and the prospects for the future looked fairly bright, even on the original lands. On this account the in- habitants were loath to go to the expense and hardships of removal to distant landS; but sold out their claims to new lands for whatever they would bring, usually only a few cents per acre. Five hundred and sixteen certificates were issued, but only 20 were located by the original claimants in person; most of the remainder were acquired by speculators in St. Louis. Perjuiy and forgery became so common that for a time a New Madrid claim was regarded as a synonym for fraud.^ PHENOMENA OF THE EARTHQUAKE. ATMOSPHEBIC PHEXOMEXA. Darkness. — ^As in most of the great earthquakes the atmosphere seems to have become darkened during the more severe shocks in the Mississippi Valley. Eliza Bryan notes that total darkness accom- panied the first shock, while a similar ''awful darkness of the atmos- phere" marked the severe shock of 4 p. m. on February 7.' Godfrey Le Sieur also says a ''dense black cloud of vapor overshadowed the land" after the severe shocks.^ At Herculaneum the atmosphere^ according to Col. Samuel Hammond, was filled with smoke or fog so that a boat could not be seen 20 paces, and houses were so shrouded as not to be visible 50 feet. The air did not clear until the middle of the day.^ A writer f rom New Madrid states that at the time of the shock the air was clear, but in five minutes it became very dark, and the darkness continued until nearly moming, during which period. there were six shocks. At 6.30 the air cleared, but at the severe shock later in the moming the darkness retumed.* 1 Carr, Luden, American Commonwealths: Missouri, Boston, 1888, p. 111. s See Bibliography, pages 111-115. • MitchilU S. L.) Trans. Lit. and Philos. Soc. New York, yol. 1, p. 201. 4 Mitchill, S. L., op. Cit., p. 297. ATMOSPHERIO PHBNOMENA. 45 The darkness was probably due to a number of cooperating causes. In all probability the dust projected into the air by the agitation of the surface, the openmg and closing of fissures in dry earth, landsUdes on dry hillsides, and possibly the fallmg of chimneys and buildings contributed to supply to the atmosphere the suspended particles which presumably produced the obscurity described. It is hkely also that aqueous vapors, rising from fissures connectmg with the warm ground waters (temperature 50° to 55°), or from the waters extruded from cracks and craterlets and Condensed by the cold December air played a part. The extrusion of such vapors, usually more or less sulphurous, is described by many witnesses. It is not entirely ünpossible that conditions f avoring condensation of atmospheric mois- ture either accompanied or resulted from the earthquake disturbance. Besides the darkness observed in the area of principal disturbance similar manif estations were recorded in other locaUties. For instance, at Columbia, Teim., a very large volume of something hke smoke was declared to have risen in the southwest, from which direction the sound appeared to have come, and, proceeding northeastward, settled as a black cloud in the course of the 10 or 15 minutes the shock lasted.^ An unusual darkness during the earthquake was reported at a num- ber of other points, but if it had any relation to the earth disturbances its nature is not known. It seems likely that in the outlying districts the darkness was due to ordinary clouds associated with storms then in progress across the countiy. ^ Odors and vapors, — Sulphurous or otherwise obnoxious odors and vapors were an attendant f eature of the earthquake at many points, as stated by nearly every writer. Bryan speaks of the complete Saturation of the atmosphere with sulphurous vapor a few minutes af ter the first shock, and of similar vapors af ter the shock of February 1? Hildreth speaks of the escape of sulphur gas through the cracks tainting the air and impregnating the water for a distance of 150 miles so it was unfit to use.^ Another observer, writing to Mitchill from New Madrid, states that although the air was clear at the time of the shock, within five minutes a vapor with a disagreeable smell and producing a difficulty of breathing impregnated the atmosphere. At Jeffersonville, Ind., warmth and smokiness were noted for several days after the shock, while at Columbia, S. C, the air during the shock feit impregnated with vapor which lasted for some time. The source of the odors in the New Madrid region seems to have been the buried organic matter which here, as elsewhere in the Missis- sippi embayment, occurs in the aJluvium and underlying Tertiaiy deposits, the emanations Coming mainly from the carbonaceous 1 Mitchill, S. L., op. dt., p. 287. > See Bibliography, pages 111-115. 46 THE NEW MADRID EABTHQTJAKE. material extruded from below through the fissures and craterletS; which were numerous in the region. In the more remote localities the yapors probably represented normal atmospheric condensations which happened to be coincident with the earthquake disturbance. ^^lÄght flaahes" and ^'glows.'* — ^The phenomena of what may be termed *'light flashes" and ''glows" seem so improbable that they would be dismissed from consideration but for the considerable num- ber of localities from which they were reported. Dillard, in speaking of the shocks (not especially the first one), says: "There issued no burning flames, but flashes such as would result from an explosion of gas, or from passing of electricity from cloud to cloud." Lewis F. Linn, United States Senator, in a letter to the chairman of the Committee on Conmierce, says the shock was accompanied *'ever and anon [by] flashes of electricity, rendering the darkness doubly terrible." * Another evidently somewhat excited observer near New Madrid thought he saw "many sparks of fire emitted from the earth/' * At St. Louis gleams and flashes of light were frequently visible around the horizon in different directions, generally ascending from the earth.* In Livingston County, according to Mr. Kiddick, the atmosphere previous to the shock of February 8 was remarkably luminous, objects being visible for considerable distances, although there was no moon. ''On this occasion the brightness was general, and did not proceed from any point or spot in the heavens. It was broad and expanded, reaching from the zenith on every side toward the horizon. It exhibited no flashes nor coruscations, but, as long as it lasted, was a diffused illumination of the atmosphere on all sides.'' At Bardstown there are reported to have been ''frequent lights during the commotions." * At Knoxville, Tenn., at the end of the first shock, ''two flashes of light, at intervals of about a minute, very much like distant lightning, " were observed.® Farther east, in North Carolina, there were reported "three large extraordinary fires in the air; one appeared in an easterly direction, one in the north, and one in the south. Their continuance was several hours; their size as large as a house on fire; the motion of the blaze was quite visible, but no sparks appeared." ® At Savannah, Ga., the first shock is said to have been preceded by a flash of light.'' That the flashes were entirely imaginary is improbable, but it is very doubtful if anything out of the ordinary actually took place. A source of many of the flashes appears to have been the thimder storms which occurred at the time. Such storms, which were very unusual at the season at which the shocks took place and which were 1 Wetmore, Alphonso, Gazetteer of the State of Missouri, St. Louis, 1837, pp. 131-142. > MItchill. S. L., Tnms. Lit. and Philos. Soo. New York, yqL 1, p. 300. * Idem, p. 288. * Idem, p. 298. K Idem, p. 287. * Idem, p. 300. V Idem, p. 28S. FISSUBES. 47 unsuspected by the people outside of their immediate path, have been recorded at a number of points, especially in South Carolina.^ Beaiing on the origin of the flashes or glows the obsenrations of several of the captams of ocean liners in the Tropica at the time of the recent severe disturbance in Mexico (1907) are of significance. They reported that on the night on which they afterwards leamed that the earthquake had occurred strong glows in the sky, resembling the anroras of northem latitudes, were seen. As these were not reported farther north the view suggests itself that they were due to magnetic dis- turbances depending upon or related to the severe earth disturbances going on at the time. It is not improbable that similar magnetic manifestations were associated with the New Madrid shock. It is probable that in the New Madrid region brush or wood fires^ made by the Indians or settlers, may have been an additional cause. It should be noted in this connection that in the New Madrid area itself, where the weather was clear at the time, no mention of any such phenomena was made in the more conservative descriptions. aEOLOGIC PHENOMENA. FISSUBES. BZCOBD OF FZssuBnra. Fissuring was one of the most common and widespread of the phenomena resulting from the New Madrid earthquake and is men- tioned in practically all contemporaneous narratives. Among the most vivid accounts is that of Le Sieur, who says ' that the earth rolled in waves several feet high with visible depressions between the swells, finaUy bm^ting and leaving parallel fissures extending in a north-south directiön for distances as great as 5 miles in some cases. Dillard' describes the opening of fissures 600 to 700 feet long and 20 to 30 feet wide. Some were sufficiently wide to swallow horses or cattle ^ and in one instance a load of Castings disappeared from a cellar.^ Another type, of smaller size but of even more general distribution, was that resulting from the settling and caving of the river banks, as described by Bradbury.* Again, Hildreth says ''the earth on the shores opened in wide fissures," and Flint states that some persons were extricated from them with difficulty,^ while Flint and others describe the f elling of trees crosswise with the trend of the fiissuring, on which persons climbed as a protection against enguU- ment. Latrobe, in relating the experience of a boatman, says ' that 1 MitehiU, S. L., op. cit., pp. 285-286. s Le Sieur, Godfrey, quoted in Am. Geologist» vol. 30, 1902, p. 80. s Foster, J. W., The Mississippi Valley, Chicago, 1869. pp. 19-22. * Flint, Timothy, Recollections of the last ten years, Boston, 1826, p. 226. * Hildreth, S. F., Original contributions to the American Pioneer, Cincinnati, 1844, pp. 34-35. 'Bradhury, John, Early westem travels, Cleveland, 1904, vol. 5, p. 205. V FUnt, Timothy, op. cit., p. 223. * Latrobe, 0. J., The rambler in North America, London, 1836, vol. 1, pp. 107-lOS. 48 THE NEW ICADBID EABTHQTJAKE. a chasm opened in the Mississippi admitting great quantities of water^ but immediately closed, giving rise to waves of great size. From many of the fissures it was stated that sand and water weie f orcibiy extruded. CSA&AOTZB OF THE nSSITBSS. Fortunately it is not necessary to rely on the contemporary accounts of the fissuring, as many of the cracks are still well preserved, and the position of others is marked by extruded materials, or occasionally by sand dikes filling the original fissures. Types. — ^There are two distinct types of what are here spoken of as fissures. The first is the simple crack, which is either open at the present time or is filled with sand extruded when f ormed or by d6bris which has fallen in subsequently. The second type, although for convenience spoken of as a fissure because of its considerable depth, appears to be in reality a narrow down-faulted block between two parallel cracks, as indicated in figure 6. The sharp depressions, often many feet in depth and width and of great length, seem to be all of this type. Plate II, Ä, shows one side of such a fissure, the drop «— *. --^^— --—- --— —-V'^- r ^ ^ Clayey all U vi um — \^ Strcccnv .•;-.*.*.*.'.*Quick*san*d.*. •.'.•. '.*.•. FiOTTRB 6.— Diagram showing trench or "fissure'* fonned by settling of fault block, a, Direction of movement of upper alluvium toward stream; h, DirectioQ of flowage of quicksand; e, Origmal walls of trench before slump. of the central segment being about 4 feet and its width (not shown) about 15 feet. In most of the simple fissures there has been no vertical movement of the walls, but occasionally low f aults are seen. No extrusion of sand has been noted from the fault block or the simple faulted fissures. Form, — ^The form varies considerably with the location and type of fissures. The simple cracks are on the whole much more irregulär than the Compound fissures. They are seldom straight for any great distance. Those near the margins of the rivers often partake of the curvature of the banks. Those across points of land in oxbows of the river are conunonly irregulär and branching. Those of the flat bottoms in sand-blow districts appear to be in most cases short, interrupted, and more or less crooked fissures, although a few are straight. The larger fault-block fissures are often straight for con- siderable distances, although usually exhibiting more or less irregu- larity of direction throughout their course. Lyell,^ in describing such fissures near New Madrid, says: "Some of them were jagged, others even and straight * * *. l might easily have mistaken them 1 Lvell, Charles, A second visit to the United States, London, 1849, p. 235. FISSUBES. 49 for artificial trenches.'* This characteristic, in fact, is very general in this type of fraetures, especially in those in the area southwest of Lake St. Francis, where the sharp steep sides and flat bottoms are especially suggestive of canal excavations. ÄrrangemerU. — The arrangement depend? upon the type and Situation. The simple fissures along river banks are usually in concentric curves. Where the deposits tended more to disintegra- tion during the shock, as on long points of land, a complex of jagged branching fissures commonly resulted. In the flat lands of the sand-blow country there appears to be a general tendency to an arrangement in parallel lines, but owing to the shortness of many of the gashes this is not always apparent. In the large fault-block fissures the parallelism is generally very marked, groups of from two to five or more long, straight, and parallel canal-like depressions being not uncommon. Intervdls. — The spacing of the fissures is very variable. They are dosest together when near the banks of rivers, parallel cracks only a foot or two apart sometimes occurring, while a spacing of from 10 to 15 feet is not uncommon. In the sand-blow districts the spacing varies from several hundred feet down to less than 10 feet, the resulting extrusions in the latter case f orming more or less confluent sheets of sand. The spacing in the caving ground on points is still less. In the case of the large fault-block fissures the spacing is greater, several hundred feet often intervening between the cracks, while the space between them may be half a mile or more. Isolated cracks of this type are not uncommon. Direction. — The direction of the fissures is of much interestbecause of the Hght it might be expected to throw upon the question of the Center of disturbance and the direction of the earth waves. In considering the direction, however, three classes of fissures must be differentiated. The first, or the simple cracks near streams, are, as already indicated, almost invariably parallel with the banks of the river«, occurring in any and aU directions, but apparently entirely independent of the direction of the earthquake waves. The second class, or the simple fissures of the sand-blow areas seem to have a tendency toward a definite arrangement along northeast-southwest lines (averaging about N. 30° E.), although where the blows are scattered the arrangement is not always very apparent. The fissures of the sand sloughs are in general parallel to the depressions and are even more commonly aligned in northeast-southwest directions. In the third class, or the Compound or fault-block fissures, the tendency is not always so manifest. In the area southwest of Lake St. Francis (T. 9 N., R. 5 E.), which was studied in most detail, the majority of the cracks have a trend very close to the bearing men- tioned, being apparently independent of the course of the river 50 THE NEW MADBID EABTHQUAKE. near which they äre located (fig. 7). There are, howeyer^ at least two Compound fissures nearly at right angles to the others. Of those who obserred the fissures soon after their formation, Flint has giyen the most specific statement of their direction, which he States was generally f rom northeast to southwest. Lyell in describ- ing those near New Madrid says:* "They varied greatly in direction, some being lO*' and others 45° west of north." FiouKE 7.— Sketch map showing earthquake features southwest of Lake St. Francis. The lines IndJcattiig flaBores are Intended to show locatton and general direction but not the specific num- ber of fhe f ractnres. It is a question if the statement credited to Lyell is not a misprint, as most other observers have recorded them as east of north. In fact, except in the case of the local fissures manifestly due to the creep or disintegration of the river banks and points, practically all the evidence supports the existence of a northeast-southwest trend of the fissuring of both simple and Compound types throughout the entire area examined. 1 Lyell, Charies, A second visit to the United States, London, 1849, p. 236. FISSUBES. 51 Length. — The length of the fissures varies considerably. Some of the narrow ones of the simple type appear to be pnly a f ew y ards long, but others may be 100 to 200 feet or even more iii length. Probably 300 feet is the maximum for this type, both when occurring along the river banks or in the broad flat sand-blow districts. The length of the Compound or fault-bloek fissures, however, is much greater. Those along the St. Francis River north of Parkin are said to be half a mile or so long (fig. 7), while others doubtless have even greater length, as the fissures cited by Le Sieur, which he reported to have a length of 5 miles. The average length for cracks of this type is estimated to be from 300 to 500 feet. East of the Missis- sippi the fissuring was in general less pronounced, but Usher ^ men- tions cracks of some size extruding for miles. Depth. — The depth of the openings was not usually very great, probably being in most cases limited to the hard clayey zone extending from the surface down to the quicksand which usually underlies the surface soil at depths of from 10 to 20 feet (fig. 6). Few openings probably extended much below the water level, w^hich is apparently nowhere much over 25 feet from the surface. The fissures locally extended somewhat deeper but did not long remain open, owing to the presence of water and quicksand. FHnt ^ speaks of the fissures as '^fearfuUy deep" at the time of his visit seven years after the shock, although formed in soft alluvium. Lyell, speaking of the trenchUke cracks near New Madrid, says ^ that his informant stated that within his recoUection these were ''as deep as wells." At the time of his visit they were only 5 or 6 feet deep, although, as he says: "The action of rains, frost, and occasional inundations, and above all, the leaves of the forest blown into them every autumn in countless numbers, have done much to fiU them up.'' The deepest fissures seen by the writer, outside the landsUde area of the Chickasaw Bluffs, were in the St. Francis River region south of the lake. Several of the Compound fissures here were so deep that when riding on horseback in their bottoms his head was below the level of the surrounding flats. Plate II, A, shows one side of one of the smaller of these fissures, the scarp being about 4 feet high. Several rather deep cracks were also seen at Marked Tree. The present depth of the simple fissures was nowhere found to be more than a foot or two, owing to their becoming filled by the caving of the sides. Of the smaller fissures, such as those of the sand-blow area, no surface indications remain other than the extruded material. FiUings (sand diJces) . — The filling of the fissures by the cß,ving of the waUs has already been mentioned. Besides this, however, should be mentioned the filHngs of intruded materials from below. 1 Usher, F. C, Am. Jour. Sei., Ist ser., vol. 31, 1837, p. 295. s Flint, Timothy, Recollections of the last 10 years, Boston, 1826, p. 223. * Lyell, Charles, A second visit to the United States, London, 1847, p. 235. 52 THE NEW MADRID EABTHQUAKE. As described elsewhere (pp. 76 to 87), the formation o£ fissures was almost always accompanied by the extrusion of sand and water which fiowed out over the surrounding surf ace. Besides the fissures which reached the surf ace, many pinched out before the top of the ground was reached, and the sand had no opportunity to escape but remained in the crack as a sand dike after the cessation of the shock. Even where the fissures reached the surface more or less sand was caught in the cracks and was lef t as stringers and dikes in the darker alluvium. (See PL III.) Probably the sand-blow region is fuU of such dikes, but owing to the f act that few cellars or other excava- tions are made, because of the f requent overflow and dampness of the region, sections showing them are rare. They are, however, rather abundant at certain points on the higher ground, especially at Charles- ton and Campbell, Mo., and vicinity, several sometimes being exposed in a single excavation. That shown in Plate III is said to be one of several in the same cellar. Ohjects swaMowed hy fissures. — Considering the number and fre- quency of fissures it at first seems surprising that so few objects were swallowed. It should be bome in mind, however, that the fissures were of two kinds, one due to the down-faulting of narrow strips of earth (fig. 6, p. 48), and the other to the cracking of the soll by tension as the earth waves progressed across the surface. The first made wide and deep depressions, which, however, had firm bottoms with nothing in the nature of deep chasms. The second were actual fissures. They were a few inches wide and extended to considerable depths, although probably rarely to the unfathomable depths which the people in their excitement were inclined to believe. The deepest one of which we have a measurement is 20 feet. The fear of being swallowed by the fissures led many to feil trees, at right angles to the direction of the fissures previously formed, upon which they climbed for safety at times of severe shocks when new filssures were to be expected. Many attributed their preservation to this practice. As a matter of fact, however, nothing except a boatload of castings, which had been stored in a cellar,^ and a few tree trunks * appear to have been swallowed in the crevices, although a few persona extricated themselves from them with difficulty. DISTBIBTTTIOir OF THE FISSTTBES. Situation. — ^The Situation of the fissures has been indicated in the foregoing discussion, but may be sunmiarized here. In general it may be said that the phenomena of fissuring may be seen on the uplands as well as the lowlands, but that in the fonner Situation they are limited to the vicinity of the edges of steep bluffs and are not seen where the surface is flat or the slopes gentle, the absence of > Foster, J. W., The Mississippi Valley, Chicago, 1860, pp. 19-22. * Lloyd, James, Lloyd's steamboat dlrectory, Cindxmatl, 1866, p. 325. EAfiTHCUAKE FISSURE FIUUED WITH IMTftUDEO SANO. CHARLESTON. MO. FISSUBES. 58 fissuring in such situations being due to ihe distance of the highland rims from the center of disturbance. Visible fissures are absent from most of the prairie ridges standing a few feet above the lowlands, because the superficial alluvial beds to which most of the fracturing was confined are there thicker than on the surrounding lowlands. In the northem part of the area the prairie deposits are more sandy and incoherent, and fissures, if they were formed at all, must have closed almost immediately. The broad, flat alluvial bottoms of Mississippi; Little, and St. Francis rivers seem to have afforded veiy favorable conditions for the development of minor fissures, for the blows marking the extru- sions from them are of wide distribution. The most favorable points, however, seem to be the broad linear depressions resulting from the sinking of the land. Here, if the extruded sand so characteristic of th^ sloughs is relied upon, the fissuring must have been even more common than on the surrounding flatlands. Fissuring in the bot- toms of bayous was also very common, and in some instances is said to have led to the draining of the waters. (See p. 54.) Locdlüies. — ^The general distribution of fissuring can best be seen by reference to the maps (PL I and fig. 1). In the figure the areas of intense and moderate earthquake action are shown. Stream banks were caved northward to Herculaneum, northeastward to Indiana, southward to the Arkansas, and probably westward throughout the Valleys of Black, Cache, and White rivers. The region showing fissuring is much smaller, being limited to the area extending from the vicinity of New Madrid to a point on the St. Francis a little northwest of Memphis. The most profound fissuring (Com- pound type) observed by the writer was along the St. Francis and Tyronza rivers in the area delineated on Plate I and shown in detail in figure 10, but it is possible that similar fissuring extends farther north, although no marked examples were seen at Blythe- ville. The minor fissuring, giving rise to the sand blows, was mainly limited to the region between a point northwest of New Madrid and the lower end of Lake St. Francis, as shown on Plate I, but similar though less extensive fissuring also occurred east of the Mississippi from southwestem Kentucky southward at least as far as the mouth of the Obion River. Fissuring also occurs near the edge of the uplands, terminating in the Chickasaw Bluffs, from near Hickman on the Mississippi to beyond the Obion River. No marked fissuring of recent date was noted in Crowley Ridge, but f aulted strata, due to early Pleistocene disturbances, were seen near Wittsburg, east of Wynne. The most northerly cracks of which evidences were noted by the writer were in the woods along the Castor River northeast of Bloom- 54 THE NEW MADBID EARTHQUAEE. field, where there are several cracks parallel to the river (one of them cutting an Indian mound) and one at right angles to the stream. Cracks were also observed parallel to a cypress bayou near La Forge, north of New Madrid. At New Madrid, which is near the northem limits of the area of marked disturbance, fissures were originally common, but many have now been obliterated by the encroachments of the river, which has cut into the banks half a mile or more since the earthquake, by the overflowing waters, by the gradual accumula- tion of silts and vegetable muck, and by the cultivation of the land. Lyell, on his visit to the region in 1849, saw several fissures still open, two of which he traced continuously for over half a mile. Some were parallel, but most of them varied greatly in direction. In depth the simple fissures ranged up to 5 or 6 feet, and in width from 2 to 4 feet, the edges being marked by accumulations of sand and dark clayey shale. He also describes a fault produced by the earthqu^ke which separated the higher level piain from the low ground about Bayou St. John just east of New Madrid. The descent was 8 or 10 feet at this point, although it was reported to be as much as 20 or 30 feet in places.^ Two parallel fissures, about 8 yards apart, through which the waters of Lake Eulalie escaped af ter the earthquake, could still be seen in the bottom of the dry bed. FiOTTKB 8.— Plan of earthquake fissare xt t> x -ii ^ -i at Beechweii, near campbeu, Mo. JNear r'ortagcville, a lew miies south- C Clayey alluvimn; 6. intruded sand ^gst of NeW Madrid, On LitÜe RivCr, laj^C cracks were observed by the writer on both sides of the river, and smaller cracks occur in the vicinity of Marston, a short distance away. At Beechweii, just northeast of Campbell, a fine section of a fissure filled with sand, pieces of lignite, and dark shaly clay, was seen in a trench. This appears to have been pushed diago- nally upward into the clayey alluvium, but not with sufficient f orce, at least on one side, to break through, the fissure terminating in a rounded head both in horizontal and vertical directions (fig. 8). Many similar fissures are found in digging cellars in the vicinity of Charleston, some distance to the northeast (PI. III), some failing to reach the surface, as in the instance described, while others reached the surface and formed sand blows. At Caruthersville, as at New Madrid, fissures were originally very numerous, but many were obliterated by the overflowing waters of the Mississippi before the construction of the levees, and by the other causes mentioned. Traces of cracks are still to be seen near the creeks and bayous. Near Carlton, between Caruthersville and Hayti, a number of faults and gashes occur south of the railroad. Most of the scars at this point seem to have an east-west direction, but one or two north-south scars occur. > LyeO, CliarleB, ▲ aeoond vMt to tbe United States, pp. 235-236. FISSXJBES. 55 Farther southwest on the Pemiscot is the site of the smoke- house incident related by Le Sieur.^ It seems that a Mr. Culberson lived on a V-shaped point in a bend of Pemiscot River, embracing about an acre of ground, on which bis well and smokehouse were situated, lying between the house and the river. On the moming of the earthquake Mrs. Culberson started to go to the smokehouse for meat, only to find the path crossed by the wide stream, the smoke- house and well being seen across the river, on the opposite side from where they were the night before. Numerous cracks were made across such pouits, some of which afforded shorter courses to the streams. Near Blytheville, east of the Pemiscot, several cracks were observed along an old bayou. Yisible fissuring occurs at Lake City, on Lake St. Francis, and it is stated that crevices can be made out in the old river banks beneath the surface of the water. It is probable that other fissures, both numerous and large, would be found if the lakes and sunk lands were drained of their waters. At Marked Tree, south of Lake St. Francis, cracks 3 to 6 feet deep and 6 to 15 feet wide, extending for some dis- tance, are found on the east side of St. Francis River, and smaller, more irregulär, and branching cra<5ks occur on the west side. One of the cracks passes through the lumber yard at Marked Tree, cutting an Indian mound. This crack is now filled with standing water during wet seasons. Similar cracks occur along St. Francis and Little rivers for 12 miles, reaching into Mississippi County (PI. I), while others are found along Tyronza River and south and southeast of Marked Tree. Cracks were seen at intervals below Marked Tree as f ar as the Maurice Smith place, in sec. 29, T. 9 N., R. 5 E., and are reported to occur near the bend of the river a mile farther to the southeast, this point being the southemmost limit of recorded fissur- ing. Figure 7 shows the distnbution of the fissures in this region. The northemmost cracks in the township are north of the Hand place, in sec. 11. These are of the simple fissure type, from 10 to 100 er more feet long and 4 to 12 inches wide, and are concentric with the curve of the river bend at that point. A mile and a half farther west, in sec. 9, a series of northeast-southwest Compound fissures of great size were observed. These are great canal-like depressions with flat bottoms and steep sides (higher even than the normal angle of repose) and so deep that a man on hoi'seback is in some places unable to see over the top. It is probable that some of the depressions are from 10 to 15 feet deep, and the bottoms are usually from 10 to 20 feet in width. One large cross fracture mainly at right angles to the others was observed, and 3 miles south in sec. 28 three or four more large Compound fissures were seen having a northeast-southwest 1 Le Sieiir, Oodirey, quoted in Am. Qeologist, vol. 90, 1902, pp. 80^1. 56 THE NEW MADBID EABTHQUAKE. trend. The Smith place fracture, aLready mentioned, was some 4 f eet deep and 20 feet across at the bottom. On the east side of the Mississippi the yisible fissures, alihough less numerous than on the west side, were not uncommon. Usher ^ describes a fissure 4 feet deep and 8 feet wide, extending sßveral miles. Lyell records several fissures in Tennessee where the ground on one side remained 2 feet higher than on the other. The writer is con- Tinced that faulting of at least 10 feet occurred at places in the vicinity of the south end of Reelf oot Lake and it is known that fissures occur in the uplands at the edge of the Chickasaw Bluffs. David Dale Owen, in his report of the geological survey in Kentucky (1854-55) states on page 119 that ''in Obion County, Tenn., depressions are even now visible 100 feet deep and varying from a few feet to iipward of 100 feet wide, whieh are said to have been more than double^ this depth when originally formed/' At another point ''in the bluffs of the Mississippi earth cracks can be traced for a quarter to half a mile^ 20 to 70 feet wide, boundcd on either side by parallel banks 1 to 5 feet above the sunk ground. " N. S. Shaler describes " earth cracks" as occurring at many points in the southwest comer of Kentucky, " the fissures being from 20 to 70 feet wide and 1 to 4 feet deep. " * To these instances the writer can add small f aults near Push and near Grace post office and a more conspicuous one parallel to Bluebank at the southem end of Reelfoot Lake. The sinking along the latter fault was about 6 feet. Besides the fissuring giving rise to visible cracks those openings from which the materials of the sand blows and sand sloughs were extruded should be mentioned. The area in which these occur is indicated in the- discussion of the blows and sloughs and its charac- teristics are considered in an earlier part of this section. CAUSE OF FISSTTBniQ. Bluff fissures. — ^In this class are included the simple fissures found near the edge of river bluffs, along the flanks of deep bayous, and near the edge of the uplands terminating in the Chickasaw Bluffs. These fissures have clearly been produced either by the actual settling of the land nearest the Channel, Valley, or other depression, the move- ment being away from the higher stationary ground, or by a partial undermining and tilting of the materials comprising the immediate banks of the streams. Where a material settling has occurred in addition to the lateral displacement, faults are produced. These fissures, as would be expected, merge into landslide disturbances. In fact, the explanation advanced appears to account satisfactorily for all the observed facts of occurrence and distribution. 1 Usher, F. C, Am. Jour. Sei., Ist ser., vol. 31, 1837, p. 295. • Shaler, N. S., Earthquakes of the westem United States^ Atlantic Monthly, November, 1860, p.666. FISSUBE8. 57 FWe» of ihe aand-hlaw regiovs.-The fissur^s from which the materials of the sand blows and sloughs were extruded were probably produced as described in the contemporary narratives. It is known that, owing to the unusually favorable conditions existing (a thin surface Stratum of stiff alluvium resting on a quicksand saturated with water as shown in fig. 6), earth waves of iinusual magnitude, sometimes several f eet in height, were propagated across the country, and there is every reason to suppose that numerous fissures resulted from this sudden flexing as they did at the time of the Charleston shock 75 years later. This " bursting of the swells, " as it was termed by the eyewitnesses, is not oniy the possible but the probable cause of the fissuring, fuUy accounting for the distribution of the cracks, for their arrangement, and for their great numbers. Fissuring of the sand sloughs. — ^The fissures of the sloughs appear to be longer than those associated with the sand blows^ but do not seem to differ in other respects. It is believed that they are due to the same processes of flexure, the principal difference being that while the surface in the sand-blow areas became flat again af ter the passage of the waves, that of the sloughs remained bent, giving rise to the troughs in which they are now found. Analogous to the slough fissures are the f aults and cracks which sometimes separate the higher from the Iower parts of the bottom lands. They are believed likewise to have resulted from warping, which, however, was on a much larger scale, the difference in level being so great and the change so sharp that fracture resulted. Compound fissures, — ^The so-called Compound fissures are believed to be the result of the dropping of a narrow fault block between two parallel simple fissures. The flatness of the bottom and the presence of trees antedating the shock seem to indicate that the present con- tour of the depression is due rather to the dropping of a block than to filling in from the sides of the fissure. There is an absence of grada- tion between the steep walls and the flat bottoms that would also be difficult to account for if a simple fissure had been partly filled by wash. In Order that the block between the parallel fissures should sink it is necessary that there should be either an outward movement of the material at one or both sides or an undermining of adequate extent. (See fig. 6, p. 48.) From examination it appears that such fissures are generally very near to rivers. The fissuring, however, is not always parallel to the stream, as would seem to be necessary if it was a lateral creep toward the river which permitted the sinking. In the majority of cracks of this type there was no extrusion of sand, hence the undermining can not be due to the transfer of material to the surface through fissures in the immediate vicinity. There re- mains the undermining by extrusion through more distant sand blows 58 THE NEW MADRID EABTHQUAKE. or undermining by creep of the quicksand into the rivers. From exposures in the bottom of the streams at low water it seems probable that the latter action took place to a considerable extent, and it is believed that in it lies the explanation of the sinking of the blocks of the great Compound fissures. FAULTS. Location. — Faulting, at least in the surficial deposits, was not a common nor a characteristic f eature of the New Madrid earthquake, although it occurred at a number of points. Some of the faults crossed the Mississippi, causing rapids and even waterf alls. William Shaler ^ in describing a patron's trip down the river at the time says that a few miles above New Madrid ''he came to a most terrific fall, which, he thinks, was at least 6 feet perpendicular, extending across the river. * * * Another fall was formed about 8 miles below the town, similar to the one above, the roaring of which he could dis- tinctly hear at New Madrid. He waited five days for the fall to wear away." Speaking of the fissures on the land in the vicinity of the town mentioned, he says many of them were ''5 or 6 feet wide, extending in length out of sight, and one side was several feet lower than the other." Lyell ' describes a fault, possibly the same as that noted by Shaler, separating the higher level piain from the low ground about Bayou St. John just east of town, the descent being 8 to 10 feet at this point. He also records several fissures in Tennessee where the ground on one side remained 2 feet higher than that on the other. At the present time evidences of faults are hard to find, owing to the modification of the bottoms by the overflowing river. The writer, however, noted a scarp along a gently sloping hillside at Beechwell, near CampbeU, Mo., the displacement of which seems to have been 10 feet or more. Small faults were also noted on the flat bottom lands near Push and Grace post offices in the vicinity of Beelfoot Lake, Tenn., and paraUel to Bluebank at the southem end of Reelfoot Lake. The displacement at the latter point was about 6 feet. Still other faults are believed to have been formed about the south end of the lake mentioned, contributing to the damming of its waters. Besides those mentioned there are a considerable number of faults parallel to a stream, where the land next the river has sunk below the landward side of the fissures. In the landsüde areas much displacement has also taken place, as is described in the aection on landslides (p. 59). Nearly all the principal fissures appear to have resulted from the dropping of nanow fault blocks, as described on pages 48 and 57. « ShiJer. wuiaim. Traas. Lit. and Philc*? S«c New Y«k, tqL 1, Pl 3QL > Lv>^ü» CtLirkä« A SKOod vssis w tbe United Sia;«s> London, ISISi, pp^ FAULTS. 59 Cause, — ^The faults appear to have resulted from various causes. Those formed across the Mississippi near New Madrid ponding the water back and giving rise to waterfalls almost certaiiily resulted from fractures incident to the uplift of the Tiptonville dorne (p. 63). Those about the southem end of Reelfoot Lake appear to be due to the same general cause. The fault described by Lyell along the Bayou St. John apparently resulted from the settling of the land nearest the river, doubtless due to undermining resulting from the extrusion of the underlying quicksands or to their lateral flowage toward the river. At Beechwell there appears to have been a general movement of the lower hillside toward the flats, leaving ä long high scarp. The f aulting parallel to the river is unquestionably due to a displacement and slump toward the river, while the dropped fault blocks appear to have been let down by undermining, as explained on page 57. LANDSLIDES. Probably no feature of the earthquake is more striking than the landslides developed in certain of the steeper bluflfs, even the casual observer noting at once the unusual character of the highly disturbed slopes. Landslides in the general sense of the term occurred when- ever a slope had sufficient steepness so that portions gave way during the shock and slipped or feil down the grade toward the lower levels. The necessary conditions were presented by both the river banks and the higher bluffs bordering the Mississippi lowlands, especially on the east side. The breaking down of the river banks was very extensive, being observed, as noted on page 53, at least as far up the Ohio as Indiana and doubtless extending down the Mississippi a corre- sponding distance. The action in the case of the bluffs was more limited in area but often far greater in magnitude. The area charac- terized by minor disturbances, such as the caving of banks, is indi- cated in figure 1, and the principal landslide area is shown on Plate I. CHICKASAW BLUFFS. Location and character. — ^The area of landslides lies along the Chick- asaw Bluffs, which face the Mississippi bottoms from the vicinity of Hickman in southwestem Kentuclg^ at least to the mouth of the Obion River, about halfway across the State of Tennessee, a dis- tance of at least 35 miles. Throughout this distance the landslides are a striking feature. Skirting the edge of the bluffs, in the vicinity of Reelfoot Lake, a characteristic landslide topography is almost constantly in sight from the carriage road which foUows their base, the road in some places crossing over large masses of earth brought down by the earthquake and winding among the trunks of the trees tilted by the 60 THE NEW MABBID EABTHQUAKE. movement. (See PI. IV, Ä.) On climbing up the bluff the traveler sees increased confusion; sharp ridges of earth altemate with deep gashes (PI. IV, B) , the whole surf ace locally being broken into a jumble of irregulär ridges, mounds, and hummocks, interspersed with trench or basinUke hollows and other more irregulär depressions, continuing with increasing frequency to the top of the bluff, along the edge of which fault scarps and fissures are of frequent occurrence. Some of the depressions, those between parallel fissures, have the same canal-hke aspect that characterizes those of the river bottoms (p. 48), but some are considerably larger, one reported by Safford * being as much as 100 feet wide. Speaking of Obion County, Tenn., Shaler says^ ''depressions are even now visible [1869] 100 feet deep and varying from a few feet to 100 feet wide." LandsUde fissures are best developed on projecting spurs, the main cracks crossing them in a direction at right angles to their axis. Where the surf aces had considerable curvature the slumping occurred in several directions, giving rise to intersecting fissures with mound- like elevations between. Although landsUdes are most strongly developed on spurs, the general slopes of the bluffs are far from being free from them, and some of the resultant mounds and troughs are of considerable mag- nitude. One of these troughs, having a depth of between 6 and 8 feet, is shown in Plate II, B. Water has collected in some of them, forming small ponds. The individual troughs are generally of con- siderable length, often several hundred feet. They begin as very slight depressions in the surf ace of the bluff, which, however, rapidly widen and deepen until strong trenches are produced. The troughs are usually not exactly parallel to the slope; the bottom descends slightly as the trough widens, and it usually has free drainage at the low end or that f arthest from its point of beginning. The individual troughs and ridges are more or less curved or irregulär. Many of them begin and end abruptly; others divide and subdivide, only to reunite again, forming a network of trenches. The movement in many cases has been such as to cause a tilting of the disturbed masses, the inclination being usually toward the bluff. The steep alluvial and dßbris fans at the mouths of the guUies extending back into the bluffs also seem to have afforded very f avorable conditions for slumping, and many of them, even at the present time, exhibit comigations and low fault scarps due to the flowage and slumping of the material toward the lowlands. Not the least conspicuous feature of the sHdes is the tilting of the older trees which, at many points, stand sharply inclined to the sur- f ace just as they were placed by the earthquake, except for a gradual i Saflord, J. M.« Qeology of Tennessee, 1869, p. 113. I Shaler, N. S., Earthquakes of the westem United States: Atlantic Monthly, Nov., 1860, p. 565. PAXJLTS. 61 recovery o£ the perpendicular by the higher growth. This feature is described in detail on a subsequent page (p. 96), and is illustrated by Plate IV, Ä. Among the localities in which landslides occurred the east side of Reelfoot Lake is the most interesting. Going south from Samburg to Push, the traveler has the landslides almost constantly in view. Some of the scars on the bluff are from 10 to 20 feet high and 200 feet long, and faults extending across the fans from the side ravines may often be seen. One and one-half miles beyond Push there is a landslide of unusually striking character, the road winding about among the mounds and ridges and beneath gigantic trees inclined as if about to fall (PL IV, A). Numerous cracks cross an adjacent fan and the bluffs above are rent into a tangle of troughs and ridges. Near Courtney Springs (Grace post office) another area of especially marked disturbance is encountered, sharp ridges formed by the slump of the hillside altemating with deep trenches in a most con- fusing manner. This locality is picturesquely called the *'Devils Hens Nest" by the inhabitants. In one of the landslide trenches is a pond, known locally as ^'Lake in the Hill,'' about 100 feet long. Originally it must have been much longer, but it has been drained by the erosion of the barrier by its outflowing waters. Cause. — The cause of the landslides was the disturbance of the somewhat delicate equilibrium existing in the bluff by the earth- quake. Geologically the bluff consists of a thick series of shaly clays extending from below the base upward for some distance, these being in tum overlain by less clayey silts, the Lafayette gravels, and finally by the loess. Water soaking downward collected in the more sandy beds above the clays, developing a sort of ''lubricating action" at the clay contact. At the same time it produced, by saturating the base of the Upper deposits, a zone of easy flowage and slipping. The bluffs, formed by undercutting at an ancient stage of the Mississippi, had not yet been brought by erosion to their normal angle of repose and were in a condition of unstable equiübrium, to destroy which only the shock of the earthquake was needed, and then down came the material in a body. This, it appears, was their mode of origin. It does not mean, however, that no landsüdes occurred before 1811; there probably were some, for it is known that small ones are still occurring. From the relative ages of the inclined and straight trees, however, it appears that a considerable part of the slipping occurred about a Century ago, and it is beüeved to have taken place mainly at the time of the New Madrid shocks. 62 THE NEW MA0BID EABTHQUAKE. WAUPINa. UPLIFTS AND DOMINO. ITie records, — Several of the contemporary accounts of the earth- quake note the occurrence of uplifts as well as sinking of the land. Thus Biyan^ states that '^the numerous large ponds which covered a large part of the country were nearly dried up. The beds of some of them were elevated above their fonner banks several feet, producing an alteration of 10 to 20 feet from the horizontal surface/' A. N. Dillard' says that ^'previous to the earthquake keelboats would come up the St. Francis River and pass into the Mississippi [through Little River] 3 miles.below New Madrid. The bayou is now dry land.'' Hildreth ' describes the heaving of the bottom of the Mis- sissippi as sufficientto tum back the waters. Flint* says "a burst- ing of the earth just below the village of New Madrid arrested the mighty stream in its course/' and states further that new islands were created. Still others record the formation of rapids by faults or doming (p. 68). Lyell,* who visited the area in 1846, says: '^From the mouth of the Ohio to that of the St. Francis, including a tract 300 miles in length, * * * was convulsed to such a degree as to create new islands in the river." In the main, however, he regards the action as one of subsidence, stating that *'there are no protuberances of upraised alluvial soil projecting above the level surface of the great piain." In this, however, he was mistaken. In 1837 Usher • described, apparently for the first time, the uplift on the south side of the New Madrid bend (Tiptonville dorne), giving a diagram showing the uplift as continuing to the mouth of the Obion River. The best account of this dome, however, is that of ' McGee,^ who bases his beHef in its existence upon the height of the banks, which are not subject to overflow, and upon the exposure of the underlying clays, which elsewhere along the Mississippi are below water level. In the course of the writer's studies numerous other elevations above the Mississippi flood plains were examined, including the long, narrow, north-south prairies shown on Plate I. Two other domelike aggregations exist. The larger of these is near Blytheville, in northeastem Mississippi County, Ark., and has a diameter of 7 to 10 miles, while the smaller is on Little River just west of Tyronza Lake and has a diameter of only 2 or 3 miles. No sections showing the lower clays were observed, but the fact that the elevations lie 1 Bryan, EUza, quoted in Am. Geologist, vol. 30, 1902, p. 78. * J. W. Foster, The Mississippi VaUey, Chicago, 1860, pp. 9-22. s Hildreth, S. F., Original contrlbutions to the American Pioneer, Cincinnati, 1844, pp. 34-3& * Flint, Timothy, Recollections of the last ten years, Boston, 1826, p. 224. <( Lyell, Charles, Frinciples of geology, vol. 1, London, 1875, pp. 452-453. « Usher, F. C, Am. Jour. Sei., Ist ser., vol. 31, 1837, p. 296. ' McGee. W J, Bull. Qeol. Soc. America, vol. 4, 1892, pp. 411-412. UPLIFTS AND DOMINO. 63 athwart the courses of the streams, in positions which apparently indicate that they can not be the result of differential erosion^ gives weight to the doming hypothesis. The three uplifts may be termed, according to the location; the Tiptonville, Blytheville, and Little River domes. TiptonviUe dorne. — ^The Tiptonville dorne, so f ar as can be deter- mined from the sections along the river bank and from the descrip- tions of Usher ^ and McGee,^ extends from the vicinity of New Madrid southward to the vicinity of Caruthersville, a distance of 15 miles, and from a point west of the Mississippi toReelfoot Lake, a distance of 5 to 8 miles. The maximum doming appears to lie in a relatively narrow belt,extendingfrom nearthesouth side of theMississippiatNew Madrid a little east of south to a point 2 or 3 miles south of Reelfoot Lake. The upHft in this belt has been sufficient to raise the surface above the reach of the highest floods. The structure of the dome is shown on the banks of the Mississippi where the river cuts the higher portion. The upper 20 f eet or so is composed of the sandy loam, clays, quicksand, etc., which everywhere characterize the alluvial deposits of tiie Mississippi, below which is exposed a thickness of several f eet of stiff bluish-green clays that neariy everywhere underUe the alluvium. These clays are not exposed elsewhere along the Mississippi, although they are encountered in numerous wells. The configuration of the dome, as pointed out by McGee,^ indicates that it was — originally a part of the broad flood plain extending from the mouth of the Ohio to the Gull, and its exceptional altitude and general confonnation suggests a localized up- lift. Moreover, several of the dry bayous enter Reelfoot Lake squarely or obliquely, and when this occurs there is no trace of delta building, and both Channel and natural levees may be traced for long distances in the lake; indeed, for some distances they may be traced throughout their extent and found to connect in the form of a fairly definite diainage System. This absence of deltas indicates that the uplift or deformation occurred suddenly. Furthermore, it is found that while great cypresses, sycamores, and poplars, sometimes two or three centuries old, grow over the general surface of the dome, no trees older than 70 or 75 years grow within the unoccupied bayous; from which it may be inferred that the uplift occurred at least 70 or 75 years ago, and prob- ably not much earlier. The amount of uplift, judging from the fact that the land is only 10 to 15 feet above high- water mark, could not have been over 15 or 20 feet. Some of the uplift was almost certainly due to the earth- quake of 1811, but if Little Prairie, on which New Madrid is situated, is a part of the dome, the upUft will have to be referred to an earUer date, as New Madrid village previous to the shock was on high land, never covered by the floods. BlyihemUe dome, — ^This dome, which, as previously stated, is 7 to 10 miles in diameter, has the same general characteristics as the !«9l>W*^" 1 Usher, F. C, loc. dt. > McGee, W J, loc. clt. 64 THE NEW MADSm EABTHQXJAKE. Tiptonville dorne, just described, with the exception ihatitscontinuity is broken by Valleys formed by the meanders of Pemiscot River and by indentations due to f onner swings of the Mississippi, the position of which is now marked by the bayous known as Clear and Fiat lakes. The elevation, if actually a dorne as there is good reason to believe, didnot result fromtheshockof 1811, for the Mississippi, which has manifestly had a part in shaping its outline, has not flowed against it since that date. The widening of the Pemiscot Valley by the meandering of the stream is likewise an earlier feature, relatively little change having taken place since 1811. The total uplift is probably not over 10 or 15 feet. LüUe River dorne. — ^Little River dorne does not differ in general character from the two just described, but has suffered probably less change in outline than the others, the oval shape being still very distinct. Little River meanders indiscriminately through high and low ground in a manner likely to result only from superimposition or from doming of the strata beneath its estabUshed bed. The depres- sion through its center seems to be an original one and not the result of efosion. In this locaUty there is no reason to doubt that the doming was the result, in part at least, of the earthquake of 1811. The uplift was probably not more than 5 or 10 feet. Oiher domes. — ^Besides the more notable domes just described there are several low elevations in the vicinity of the sunk lands which may be of similar character. Near the outlet of Big Lake, in Mississippi County, Ark., just south of the Missouri line, especially, the topog- raphy seems to indicate that the subsidence of the lake basin was accompanied by a corresponding elevation inmiediately to the west, which shut off to a certain extent the original drahiage, and gave rise to the present lake.^ Cause of uplifts and doming. — ^The hypothesis that the writer regards with most f avor as an explanation of the uplifts and doming, although it must be held as largely lacldng in definite proof , assumes a lateral movement of the deposits of the central part of the embay- ment area toward a fault having the direction and position indicated by Plate I. This movement took place by virtue of a subsidence of the ''Mississippi trough,'' and formed the corrugations represented by the domes and basins shown in figure 18 (p. 108). The matter is further discussed on pages 105 to 109. DEPRESSION — "SUNK LANDS.'* No other feature of the New Madrid region is so conspicuous and striking or so widely known as the so-called ''sunk lands," resulting from the local settling or warping of the alluvial deposits of westem i Morgan, A. E., and Baxter, O. G., Report on the St Francis Valleydrainage pro^t; BnlL Offio^ Esper. Sta. No. 230, U. S. Dept. Agr., pt. 1, p. 11. F SHALLOWER Li (< «»»^^^ * . ^^^« 9 9 DEPRESSION — SUNK LANDS. 65 Tennessee, southeastem Missouri, and nortlieastem Arkansas by the action of the earthquake of 1811. The accounts of eyewitnesses are particularly vivid and teil not only of the major alterations, which can stiQ be recognized after the lapse of a hundred years, but of many minor changes now difiicult to distiaguish from the normal irregu- larities of the Mississippi bottoms. TYPES. The sunk lands may for convenience be divided into (1) those marked by sand sloughs, (2) those characterized by river swamps, and (3) those covered with lakes of standhig water. The sand sloughs, which are elsewhere described (p. 83), are broad, shallow troughs, generally of considerable length, several f eet in depth, and marked by more or less extensive, well-defined ridges covered by extruded sand interspersed with depressions in which the timber has been killed by standing water (PI. IV, B) . The river swamps include the depressed areas along certain of the streams (fig. 9), the level of which is such that water Stands over them for considerable periods, but does not cover them so deep as to prevent the growth of timber. They are therefore characterized by wet-land timber, most of which is young growth, although it may Original level of crest of stream bank I ffvnl nf ntrnnm hr tainly the water Stands from 6 to 10 feet deep over the original flood 68 THE NEW MADBID EABTHQUAKE. plain at many points or f rom 20 to 30 feet or more over the bottom of the original Channel. The deeper parts of Reelfoot Lake, including the Channels of the old Keelf oot and Bayou du Chien creeks and the depressions known as basins (fig. 12), are reported to be from 14 to 20 feet deep at low water. It is probable that in this lake the water Stands from 2 to 10 feet, or possibly even more locally, over the original land surface. DISTRIBX7TI0V. The known sunk lands are practically limited to the flat bottom lands of Mississippi, Little, and St. Francis rivers, and those on each of these streams are f ound between structural or topographic ridges of greater elevation; the eastemmost between the CUckasaw Bluffs and the Tiptonvüle-Blytheville line of doming, the second between the dome mentioned and the prairie ridge extending from Big Lake to beyond Maiden, and the third between the prairie indicated and Crowley Ridge. Sinking doubtless occurred at many other points, as, for instance, at New Madrid, but in general its occurrence can now be recognized only on the bottom lands where the obstruction of drainage makes it apparent. The general distribution of the sunk-land areas between the struc- tural and topographic ridges has been indicated in the precedbig Paragraph and it only remains to present a few descriptive notes on the occurrences at various points. On Flate I are shown a considerable number of sunk-land areas in the St. Francis basin, while a smaller number between Campbell and Cairo are also indicated. The sinking in the latter region was much less than in the region south of Campbell and New Madrid, but the fact that the areas indicated occur within the region of marked dis- turbance, as evidenced by the numerous sand blows and by the fissures, the latter of which occur at least as far as Charleston (PL III), and by agreement in general characteristics with the known sunk lands (fig. 10) rather than with the swamps of the bayou type (fig. 11), leads the writer to class them as sunk lands. In general, however, there is far less water than ia the southem sunk lands, and little timber was killed by submergence. The most northem locaUty at which conspicuous examples of sunk lands have been reported is in the vicinity of New Madrid. Hildreth ^ states that the site of the town was origiaally 15 or 20 feet above the fioods of the river, but sank so low at the time of the earthquake that the next rise covered it to a depth of 5 feet. According to Bryan ^ the settling amounted to 15 feet, although half a müe below town there was no change in the banks whatever. Bradbury ' says that i Hfldreth, S. F., Origbutl contributions to the American Pioneer, Cincinnati, 1844, p. 35. » Bryan, Eliza, quoted in Am. Geologist, vol. 30, pp. 77-78. » Bradbury, John, Early western travels, Cleveland, 1904, vol. 5, p. 210. DEPRESSION — " STTNK LANDS." 69 Z Mlles Brue obstruction to Rceifoot Creek FiGUBB 12.— Map of Beelloot Lake and its submerged dralziage Channels. 70 THE KBW MADHm KABTHQtTAKE. the greater pari of the plain on which the town was situated became alake. Lyell, who yisited the region in 1846, when the evidences were much clearer than at present, says of the district west of New Madrid: The Lugest area affected by the convulaion lies 8 or 10 miles westward of the Mis- siasippi and inland from the town of New Madrid, in Missouri. It is called the "sunk country" and is said to extend along the course of the White Water (LitÜe River?) and its tributaries for a distance of between 70 and 80 miles north and south and 30 miles or more east and west. Throu^out this area innumerable submerged trees, some Standing leafless, others prostrate, are seen; and so great is the extent of the lake and marsh that an active trade in the skins of muskrats, minks, otters, and other wild animals is now cairied on there. In March, 1846, I skirted the borders of the sunk country nearest to New Madrid, passing along the Bayou St. John and Little Prairie, where dead trees of various kinds — some erect in the water, others iallen and strewed in dense masses over the bottom, in the shallows, and near the shore — ^were con- spicuous.^ Farther south similar conditions exist. Dillard ^ says: I have trapped there [in the region of the St. Francis] for 30 years. There is a great deal of simken land cansed by the earthquake of 1811. There are large trees of walnut, white oak, and mulberry, such as grow on high land, which are now seen submerged 10 and 20 feet beneath the water. In some of the lakes I have seen cypresses so far beneath the surface that with a canoe I have paddled among the branches. The most northem point at which these sunk lands west of the Mississippi were examined in detail by the writer was near Kennett, Mo., where a trip was made by dug-out on Vamy River, a tributary bayou of the St. Francis. Along this river are considerable Stretches of water destitute of timber, except for an occasional dead or nearly dead cypress, but covered with a dense growth of water weed, through which a narrow but deep Channel meanders (PL V, Ä). Working the boat through the tangle of weeds, it is found that the water about the gaunt dead trunks (PI. IV, B) is many feet deep. Nothing can be seen of the enlarged butts which normally characterized the cjrpress trunks from the water surface upward for several feet, but on sound- ing with an oar the enlargements may be detected. As the water was, at the time of the writer's visit, at its usual height, it is apparent that there must have been a rise in its normal level of 5 to 8 feet or more since the trees were young. In fact, the original banks of the river appear to be submerged several feet below^ the water, even in dry seasons, while the average submergence may be as much as 10 or 15 feet. The greater part of the timber in Varny River sunk lands has long since fallen and been covered by the waters, though some dead or nearly dead trees (PI. V, A) remain. The latter apparently stand in relatively shallow spots where the submergence was not sufficient to kill the cypress immediately, the trees lingering 1 Lyell, Charles, Prmciples of geology, vol. 1, London, 1875, p. 253. * Fester, J. W., The Mississippi Valley, Chicago, 1809, p. 21. Q GROWTH ALOHQ E DEPRESSION — SUNK LANDS. 71 for years, some even to the present time, before succumbing to their changed environmeixt. Along the edges of the submerged lands the young trees form a dense wall of timber (PL VI, Ä), invariably characterized by enlarged butts and contrasting strongly with the straight trunks of submerged trees (PL IV, B). This young growth has Sprung up since the shock and is rapidly encroaching on the swamp (fig. 13). The sharp Kne of demarcation is shown in Plate VI, B, in which the young growth to the right is contrasted to the open portion in which the timber was killed to the left. On nearer approach the details shown by Plate V, B and VI, A are observed, the enlargements of the butts above the water being especially notice- able in contrast with the straightness of the older trunks at the water's edge. Apparent evidences of faulting in this region have been cited by Prof. E. M. Shepard, who, describing a trip made several years ago into the swamps, states that *'to the east of a north and south line at one point in the swamp the timber grows tall and ereet, while to the ^ubmerged Present water level , n cypress ^ ..^..Fprm?D!:r^'^ftter |eyej _ FiGURE 13. — Section showing cypress killed by submergence and yoang growth springing up in the less deeply submerged areas. The present water level shown is that of normal high water. West it appears to be submerged to a depth of from 10 to 15 feet.'** The present writer was unable to verify the occurrence of this sup- posed fault, and no such features were noted by him in any of the other lakes or sunk-land swamps examined. Near Caruthersville there was considerable local sinking and a number of small sunk areas occur, but many were filled or rendered unrecognizable by the overflow of the Mississippi before the construc- tion of the levees. Walnut stumps are said to be found just below the surface in swamps along the railroad. In the *'Lake Nicormy swamps,'' a few miles west of Caruthers- ville, the conditions are said to be similar to those near Vamy River, but the submergence appears to be less and more of the old timber has survived, especially near the edges. Big Lake, located in Mississippi County, Ark., near the Missouri line and south of Vamy River, presents all the usual evidences of forest and other features characteristic of sünk lands submergence. The bottom of the lake is covered with a fallen forest of hardwood of varieties that Commonly grow on dry ground, the prostrate trunks of which nearly all he in the same direction. According to the engineers 1 Shepard, E. M., The New Madrid earthquake: Jour. Geology, vol. 13, p. 50. 72 THE NEW MADRID EARTHQUAKE. in Charge of the St. Francis Valley drainage project, this lake is the most obvious example of sunk lands in eastem Arkansas^ although the ponding of the water was probably in part due to an upUft crossing the old drainage lines.^ On the Hatchie Coon Sunk Lands, stretching along the St. Francis in Craighead County, Ark., oak, cypress, sweet gum, etc., are found Standing in water up to their branches and in some places buried in sand, etc. ; it is said that the roots are 50 feet below the surf ace. All are dead. Elsewhere, especially near Lake City, considerable ex- panses of water entirely open except for the water weed were noted, the original timber having fallen long ago. Where the water is shal- lower, however, great trunks, most of them lifeless, still rise to con- siderable heights, indicating that many of them must have survived long after the submergence due to the earthquake. Plate IV, 5, shows two of these trunks rising straight from the water with no sign of the enlarged butts, which, however, were detected with an oar a number of feet below the water surface. The encroachment of the young growth is not so conspicuous as on Vamy River. The original flood piain Ues deep beneath the waters, except in periods of extreme drought, when the waters subside so that they barely show at the surface. At a similar stage of water outside the sunk area the trunks would be normally many feet above the water level. Li Lake St. Francis, a continuation of the enlargement of St. Francis River, just south of the Hatchie Coon Sunk Lands, the con- ditions are very similar to those just described. The growth, so f ar as it exists in the lake, is reported to be young, consisting of wet-land varieties which have sprung up since the high-land species were killed by the submergence. Old cypress stumps and submerged trees pro- jecting from the water are said to be found for a distance of one-half to 1 mile from shore. Submerged trees over which one may float in a canoe are reported in Golden Lake, near Wilson, Mississippi County, Ark., and in other lakes in that vicinity. In Lake Tyronza submerged timber is also common, as it is along Tyronza River for several miles northeast of Tyronza village. Dead Timber Lake, a depressed stream channel connected with Tyronza River, is an open lake with water 6 to 8 feet deep from Tyronza to Deckerville. At times of low;^ water stumps as numerous as the trees in the thickest forest are said to be visible along its bed. In addition should be mentioned Flag Prong and Sunk Land Lake, secs. 3 and 4, T. 9 N., R. 5 E. (fig. 7). The former is a straight sunk trough about half a mile long, in which stand submerged stumps of mulberry and other highland species. The new growth is mainly 1 Morgan, A. £., and Baxter, O. O., Rept. on the St. Francis drainage project: Bull. Office Ezper. Sta., U. S. Dept. Agr., No. 230, PI. I, p. 10. B. ELEVATED CVPRESS MEAR SOUTH ENO OF REELFOOT L DEPRESSION — SÜNK LANDS. 73 flags. In Sunk Land Lake the conditions are essentially the same. Log Lake seems to be a depressed bayou as it is much wider than the ordinary Channel of the St. IVancis and too small to be a Mississippi bayou. No sunk timber is reported in it, but Uving cypress trees hundreds of years old stand in its bottom. If it is to be regarded as sunk land it must be referred to an earthquake long antedating the New Madrid shocks of 18 IL Willow Pond, in sec. 6 of the same township, was marked as sunk land by the land survey of 1845-46, but no stumps are now visible. The bottom is very sandy and Sup- ports only flags and willows. The absence of cypress timber, which must have obtained a foothold if the pond was much oldör than 1811, suggests that notwithstanding the absence of stumps it is to be classed with the sunk lands produced by the New Madrid shock. The most typical sunk land of the whole earthquake area is Reel- foot Lake in Tennessee. This lake is 8 or 10 miles in length and 2 or 3 in breadth (early Statements gave it lengths up to 75 miles and widths of 5 to 10 miles). McGee^ in a description of the lake says: Here and there, particularly toward the westem side, groves of sickly cypresses spring from its bottom and half shadow the water surface with puny branches and scant foliage, and here and there throughout all portions of the water body, save in the Channels of the old bayous, gaunt cypress tninks with decaying branches stand, sometimes a dozen to the acre, numbering many thousands in all. Moreover, be- tween the decaying boles, rising a score to a hundred feet above the water, there are tan times as many stumps, commonly of lesser trees, rising barely to low- water level. Reference to figure 12, compiled from Information coUected by the writer, will bring out clearly the shape and general conditions at the lake. On it will be seen the positions of the submerged Channels of Bayou du Chien and Reelfoot Creek, at the junction of which an old sycamore stump, many feet in diameter, once standing on high ground, and marking a comer of one of the old Spanish land grants, can still be seen. Here and there are the basins, probably repre- senting, as McGee suggests, old Mississippi bayous orthe"lakes" of the inhabitants. The submeigence in Reelfoot Lake ranges from 5 to perhaps 20 feet, although greater depths are reported. The submerged area was well wooded, much of it being covered with species characteristic of dry situations. Most of the timber remained upright after the shock but was gradually killed by the rising waters. The trunks retained their branches for a time, but these gradually dropped away. The better timber was cut off, as shown by the lower stumps in Plate VII, Aj but the poorer trees were allowed to stand until they finally broke oflF, leaving the shattered stumps shown by the view. Through- out practically the whole extent of the lake stumps may be seen projecting above the water, or the remains of old logs may be detected 1 McGee, W J, A fossil earthquake: Bull. Geol. Soc. America, vol. 4. 1892. 74 THE KBW UADBID SABTHQÜAKB. beoeath the sutfac«. Orer large areas the tnmka are upnght, pro- jecting above the surface bj thousands as shown in Flate Vü, A, from a view taken near Sambui^. The visible stumps, however, represeat but a small part of the whole, thousands having rotted or broken off below the aurface. Many are chaired as if bumed off. Owen,' in fact, says that fires frequently swept from the shores outward through the closely crowded dead timber of the lake. So thick and hi^ was this dead timber as recently as 1855 that Safford* stated that even where the lake could not be seea its outline could be traced when looking out over the forest by the lines of dead cypress around ita borders. As in the case of the Missouri and Arkansas swamps the difference between the old and young growth is distinct and immistakable, old straight-sided submerged tnmks being fre- quently Seen aide by side with a younger growth with the character- iaüc enlarged butts. Near the submeiged Channels of Keelfoot and Bayou du Chien creeks the stumps are indined (fig. 14), thür slant being appar- ottly due to the slumping of the bauks of the natural levees from wiiich they grew. In other places the timber over la^e areas was completely proatrated, and where the water is shallow hundreds of tnmks may be aeen lying below the surface or harelj projecting aboTe it. Kew growth is starting in the shallow parts of the lake, especially «long the old natural levees, vhich are relatirely near the surface. Along Reelfoot Creek the land was low and swampy before the earthquake, and cypress consütuted the predominating timber. On the higher ground, however, walnut. ash, oak, elm, catalpa, mulberty, etc., abounded, and their stumps mar still be rec<^nized among those in the lake. c&DsB OF smove. The sunk lands may be divideo«-iated with extrasion of sand and mav bare reaulted from DEPRESSION — SUNK LANDS. 75 undermining incident to the ejection of the qmcksand beneath the Upper alluvial beds or from the contraction of this sand caused by its agitation and readjustment by the earthquake vibrations and the loss of its water. The more extensive sunk lands, however, especially when taken collectively (as the St. Francis group), seem to require a more potent cause, there being far too little extruded matter to account for the depressions. This cause is presumably allied to those producing the doming, which are considered in connection with that phenomenon (pp. 64 and 106). In brief, the sunk lands probably resulted from a warping and lateral movement incident to a sub- sidence in the floor of the basin, approximately along the line of the faidting that gave rise to the earthquake. In the formation of Reelfoot Lake, the best known of the sunk- land areas, there appear to have been several contributing causes. Several of the earlier writers mention the extrusion of sand along the lower part of Reelfoot Creek in amounts sufficient to clog the Channel and obstruct the drainage, and this may have assisted in ponding the waters. The land at the south or lower end of the lake, however, now Stands from 10 to 25 feet above the level of the lake bottom, although both appear to have been originally a part of the same approximately level surface. Both sinking and uphft seem to have taken place. The land about Reelfoot Lake has undoubtedly sunk, giving an increased gradient and acclerated velocity to the little streams entering from the east, with the result that the old Valleys and f ans have been trenched in many places by small but sharp Channels. On the other band, the bottom of Reelfoot Creek Channel is a number of feet higher in what was originally the lower part of its course than in the old upper part now covered by the lake. Again, south of Reelfoot Lake cypresses with fully developed knees, such as develop only in water, are foimd on the high land not now flooded, pointing to a differential elevation of this region. The main move- ment along the lake appears to have been downward and was due in part to a general depression and in part to faulting. Several of the faults are mentioned on page 58. The uplift at the foot of the lake seems to have formed a part of the general uplift of the Tiptonville dorne. The line between the uplifted and submerged lands at the south end of the lake is so sharp that it suggests faulting or at least a very sharp flexure. West of the Mississippi the conditions were essentially the same, for although, as in the Reelfoot region, the land was low and wet before 1811-12, the waters were much deeper after the earthquake, the ponding being a result of subsidence and the formation of basins combined with a certain amoimt of drainage obstruction due to local elevation. 76 !rHE NEW MADRID EABTHQUAKE. EXTBTJSION. RECORDS. That the phenomena of extrusion, including the ejection of water, sand, mud, and gas, were among the most noticeable features of the earthquake is shown by all published accounts, and this has been fully substantiated by the writer's studies of the extrusion deposits represented by the sand blows and sloughs. Probably the most reliable account is that of Bringier, the engineer, who sajTS * that the water forced its way through the surf ace deposits — blowing up the eaxth with loud exploeions. It rushed out in all quarters, bringing with it an enormous quantity of carbonized wood, reduced moetly into dust, which was ejected to the height of from 10 to 15 feet, and feil in a black shower, mixed with the sand, which its rapid motion had forced along; at the same time the roaring and whistling produced by the impetuosity of the air escaping from its confinement seemed to increase the horrible disorder * * *. In the meantime the snrface was sinking and a black liquid was rising to the belly of my horse. Bradbury' records a case reported to him by an observer of a chasm which on closing threw water to the height of a tall tree. Similar Statements are made by Hildreth,* Foster,* and Lloyd,^ but are apparently based on current reports rather than on Observation. The quantity of water extruded was enormous. In the vicinify of little Prairie (Caruthersville), according to Flint,® the amount was sufficient to cover a tract many miles in extent from 3 to 4 feet deep. Some districts were still covered at the time of his visit seven years later. Several writers record the extrusion of gas besides the sand, mud, and water. Hildreth ^ says : The Bulphurated gases that were discharged during the shocks tainted the air with their noxious effluvia and so strongly impregnated the water of the river to the distance of 150 miles below that it could hardly be used for any purpose for a number of days. CHARACTER OF EJECTED MATERIAL. As to the nature of the ejected material there are several lines of evidence. On one hand its character is mentioned in nearly all accoimts of the disturbance and careful contemporary examinations were made and recorded, while on the other we have at the present time evidences of its nature in the sand blows, sand dikes, and in the material brought up by wells and Springs. The material ejected everywhere consisted of sand, with a certain admixture of carbona- 1 Bringier, L., Am. Jonr. Sei., Ist ser., vol. 3, 1821, pp. 15-40. « Bradbury, John, Early westem travels, Cleveland, 1904, vol. 5, p. 209. * Hildreth, S. P., Original contributions to the American Pioneer, Cincinnati, 1844, pp. 34-35. « Foster, J. W., The Mississippi Valley, Chicago, 1869, pp. 19-25. » Lloyd, J. T., Lloyd's steamboat directory, Cincinnati, 1856, p. 325. • Flint, Timothy, RecoUections of the last ten years, Boston, 1826, p. 222. » Hildreth, S. P., op. cit., p. 35. . SAND CONTAIHING LISMITE P*RTICLES FROM DEEP WELL AT MEMPHIS, TENN. EXTRUSION. 77 ceous material, and an occasional foreign fragment; such as the cranium of the extinct musk ox (Bootherium hombifrons) later donated to the Lyceum of Natural History of New York. The sand; which always constituted by far the greater part of the ejected material, consisted, as testified by the sand blows and crater- lets, mostly of quartz with a certam admixture of clay and of Ugnitic and other organic material, which possibly gave rise to the ^'sulphur- ous'' odors described. Afterward the clay and carbonaceous mate- rial was washed out and the sand became bleached neariy white, forming the characteristic white patches and belts that contrast so strongly with the ordinary black soil of the bottoms. The sand seems to have ranged from coarse to fine, a medium grain predominating. Most of the contemporary accounts speak of the carbonaceous material as "coal,'' but others speak of it as "carbonized wood'^ or lignite. The material seen by Lyell near New Madrid is described in one place as bituminous coaly shale (clay), such as outcrops in the river bank and is found in shallow wells 35 feet or so below the surface, and in another as lignite. The best description of its behav- ior on combustion is given by Mitchili, who examined samples sub- mitted by a correspondent.^ I found it very inflammable; it consumed with a bright and vivid blaze. A copious smoke was emitted from it, whoee smell was not at all sulphurous, but bituminous in a high d^ree. Taken out of the fire in its ignited and buming state, it did not imme- diately become extinct, but continued to bum until it was consumed. While blowed upon, instead of being deadened it became brighter by the blast. The ashes formed during the combustion were of a whitish color and when put into water imparted to it the quality of tuming to a green the blue coroUa of a phlox whose juice was sub- jected to its action. Some specimens of the hgnitic matter were coated with a whitish or yellowish substance, suggesting sulphur, but it was probably the sulphate of iron common in hgnite and certain coals. Wood not lig- nitized was also reported by some observers. The largest pieces of lignite observed by the writer were from ^ inch to 1 inch in length and were obtained from the sand dike at Beachwell near Campbell, described on page 54 and illustrated in figure 8. More commonly the lignite occurs as minute particles mixed with the sand grains, as illustrated by Plate VIII, A, which shows the appearance of Tertiary sands from a well at Memphis, as they appear when greatly enlarged under a microscope. The samples were fumished by Prof. E. M. Shepard. TEMPERATURE OF EJECTED WATER. Most witnesses say nothing of the temperature of the ejected waters, but positive statements are made by one or two. For ' Mitchill S. L., Trans. Lit. and Fhilos. Soc. New York, vol. 1, p. 298. 78 THE nSW MADKID KABTHQUAKB. instance, both Godfrey Lesieur ' and & correspondent of Sfitchill ' report lukewann water extnided from the cracks in the New Madrid area. The extrusion ot warm water ia also reported by Haywood,' This presumably arisee from the subterranean warmth at even mod- erate depths. The mean annual temperature of the region is about 54", which would be spproximately the temperature of the groimd- water st the depth of no Taiiation, say 50 feet below the surface. Water of Üäa temperature emerging on a cold December night would be ahnost sure to f eel warm. If it came from any considerable depth the temperature would be about 1° higher for every additional 50 feet of the depth of ite source. ESCAPE OF GAS ANI> WATER AFTEK THE EABTHQUAKE. In a few places the ejection of water and gas continued for some time af ter the cessation of the earthquake shocks. Fester in 1869 * notes that "blast« of air and gas yet found their way to the surface through many of the half-filled fissures," and more recently Shepard has maintained that subterranean waters in the form of Springs are still being forced up. He says: A coreful study of theae fstxeamB, eepecially along the St. FibucIh, the Little Tynmzfl, and the Big Bay reveale the fact of the conatant escape of vater from smaJl openisga flUiTounded by little conee of eand. Thia ia noticeable for miles along the St. Francia and in the Big Bay diatrict, eepecially on the bluff aide of the streams. Deep-eeated wat«r, then, ia constantly Coming to the suiface, brisgiiig vith it fine Band from below. TnvU IS.— IH>gl*mihawliig normal mDvemenUofgroaDdinMinidiMriveTtatMmillBpriiigHBlotit The wxiter visited the region in Company with Prof. Shepard, but reached altogether difFerent conclusions, The "small openings sur- rounded by little cones of sand" are minute Springs from 1 to 4 inches in diameter, which, aa the writer's hydrologic studies lead him to con- clude, differ in no way from the Springs that everywhere rise in the bottom of Btreams under similar conditions. The volume of a single spring is only a fraction of a gallon a minute, and the whole flow of the streams depending upon the Springs for their supply was well > L«8lear, Oodfrey, quoted In Am. Oeologlst, toI. 30, 1902, pp. T9-S0. ■HiUlilU,8.L.,1oc.dt. • HaywDod, lolm, The natnisl and Bboriginal hiBlorr ol Tenneasee, NashTlUe, ISS, iDtroductoiy chaptcr. i> ♦. Where the white sand was blown up, it lay aroimd the hole in a circular form * * * hillocks of white sand of the size of potato hills. These are all through the Ghickasaw country. In the vicinity of Ghickasaw Bluffs, according to Safford,' ''sand, fine gravel, and fragments of lignite Ccoal') were 'blown up' through the fissures and are now found in little ridges or hillocks." Data coUected by the writer leads to the behef that they were found southward nearly to the mouth of the Obion River. SAVD 8I.OTrOHS. Nature. — Sand sloughs (the word is mvariably pronounced ''sloos" in the New Madrid region) are linear depressions usually 3 to 5 feet, occasionally more, below the general level of the bottom lands. They are normaJly characterized by low somewhat ill-defined ridges of sand parallel to one another and to the trend of the depres- sion altemating with shallow troughs, and they collect water into long narrow pools, the largest and deepest of them giving rise to ponds or lakes. The ridges and depressions are as a rule only a few inches in height or depth, but they have had a marked effect on the timber growth. Thus many of the sandy ridges practically refuse to Support Vegetation and open strips' tend to result as the old trees die without their places being taken by yoimger growth. In the pools much of the old timber, which was largely of highland varie- ties, was killed. The stumps are still visible in many places, but as yet a new growth has f ailed to obtain a f oothold, although the ponds are often surrounded by willows. The sand sloughs are not infre- quently a mile in width, and may be considerably greater. They grade into what may be termed "sand scatters" in which the whole surf ace of the higher as well as the lower portions of the bottoms is covered with a thin sheet of sand. The material is the same as that of the blows, being a medium white quartz sand, containing a few particles of lignite or chips of a hard shaly clay, Situation. — The sand sloughs are limited even more strictly than the blows to the lower lands, apparently representing the concen- tration of the phenomena of which the blows are isolated units. Since depressions are necessary for the production of the phenomena, > Haywood, John, The natural and aboriginal history of Tennessee, Nashyille,1823,introductory chapter, > Saflord, J. M., Geology of Tennessee, NashTÜle, 1800, p. 113. 84 THB STEW ICADSm EABTHQTJAKE. it foUows that the sloughs are always foimd on the Iowest ground and are most conspicuous in the broad, flat bottoms nearest the uplands, rarely near the Mississippi. Distributian. — The sand sloughs and allied phenomena occur in three main belts, the most extensive being in the St. Francis Valley between Crowley Ridge on the west and the prairie ridge extending from Big Lake on the south to beyond Maiden on the north. This area is shown on the general map ^1. 1), and in more detail in figure 10. A second belt ahnost as extensive, but less strongly developed, lies east of the prairie ridge mentioned and west of the uplif t marked by the Tiptonville, Blytheville, and little River domes. The third belt lies between the Tiptonville dome and the Chickasaw Bluffs, mainly in the Reelfoot Lake region. The first may be called the St. Francis, the second the Little River, and the third the Reelfoot area. In the St. Francis area the slough extending southwest from Lake St. Francis to the Big Bay at the St. Francis (PL I) was examined in 1904. It is a broad sandy belt, some 2 miles in width, occupjdng a low depression in the bottom lands, marked by numerous long narrow pools, often several to a mile, and characterized by stumps of dead timber. Near Marked Tree several sand sloughs, with the timber either wanting or of small size, are found, but the absence of old stumps suggests that the depression occurred before 1811. The scarcity of young timber is due to the continued presence of the water. Unfortunately, owing to the high water existing during the more detailed field work in the spring of 1905, the St. Francis sloughs f arther north were entirely submerged, and no opportunity f or their detailed examination presented itself . From the Information that could be coUected, however, it appears probable that the sloughs are of much the same character as in the southem part of the area, but with a general increase in the amount of the sand to the north. The sloughs, which are generally associated with the belt of sunk land swamps, are, as shown by the map, much more numerous at the northem end of the area. In the Little River area, sand sloughs occur in the Big Lake and Lake Nicormy districts, and in the East swamp region east of the West and Rosebriar prairies near Maiden. As in the St. Francis area the amount of sand increases toward the north. In the Reelfoot district, sand probably resulting from extrusion is said to be found at several points beneath the water of the lake, most of them away from the beds of the submerged creeks. Sand ridges several feet in height and surrounded by muddy bottoms are reported by fbshermen. At times of flood strong currents from the Mississippi River flow through the lake, however, and are known to have scoured great quantities of sand from the lower end and spread H DOUBLE SETS OF ROOTS AT SOUTH ENO OF BEELFOOT LAKE, TENN. CAUSE OF EXTBUSION. 85 it over the flats to the south, forming the so-called scatters^ and it is not impossible that the submerged ridges mentioned are due to the same general cause. There was, howeyer, a considerable extrusion of sand outside of the submerged area, and local sand sloughs doubt- less occur. SAVD 80ATTE&S. The term ''sand scatters" is here applied to the thin surface cover- ing of sand, which can not be differentiated into individual blows or sloughs. The scatters are most commonly developed on low ground, but are also found at higher levels. In composition they are similar to the blows and sloughs, but generally contain less lignite. They are probably of diverse origin, some probably representing confluent blows, others expanded sloughs, and still others the recent wash from the streams. The scatters leading southward from the foot of Keelfoot Lake are of the latter type. In the northem part of the area the so-called scatters are of ten only sandy phases of the alluvium. This seems to be true in general of the part of the earthquake area north of New Madrid. TheKeelfoot scatters are the most conspicuous and are still in active process of construction or modification. Leading southward from the ''Washout" at the extreme southem end of the lake, a broad belt of sand, with occasional pebbles of bronzed gravel resembling the Lafayette, extends southwestward to the Mississippi below Caruthersville, a distance of about 15 miles. Near the lake the sand is in places from 5 to 8 feet in thickness, and has so deeply buried the trees as to destroy them, the dead stimips now projecting from the sand (PI. X, Ä). At other points, where the sand bed was only 2 or 3 feet thick, roots have been sent out near the new surface and the trees have survived. In some places a shifting of the currents has removed the sand, showing the trees with two sets of roots, one at the present surface and one at the former level of the sand (PI. X, B) . Over much of the area the sand is destitute of Vegetation, harren tracts several acres in extent, marked by large wind ripples and drifting dunes, being not uncommon. The sand of the scatters has come largely from the ''Washout" at the south end of the lake. This is a sharp trench from 15 to 40 feet deep, one-f ourth of a mile wide, and a mile or two in length. Great as the trench is, however, it does not seem competent to have fur- nished all the material of the scatters, and it is suggested that some of the sand may have been scoured from the lake bottom, leaving certain of the supposed ridges of extruded sand. CAUSE OF EXTBUSION. The extrusion appears to be of two general types, the first of which may be characterized as violent ejections and the second as 86 XHB NEW UAX>Bm EAETHQUAKE. quiet extrusion. To the first type belongs the jets described by Bringier,* Le Sieur,' and other contemporary writers, and to the second the more slow and quiet extrusion giving rise to the sand blows and sloughs. In regard to the cause^ Bringier says: The violence of the earthquake having disturbed the earthy strata impending over the subterraneouB cavities, existing probably in an extensive bed of wood, highly carbonized, occasioned the whole superior maas to settle. This, preeaing with all its weight upon the water that filled the lower cavities, occaaioned a displacement of this fluid, which forced its passage through, blowing up the earth with loud explosions. Le Sieur, on the other band, ascribes the ejection to the bursting of the swells or earth waves which progressed across the surface. Bradbury records extrusions due to the sudden closing of fissures,^ and several other writers are inclined to similar views. Of the later writers, Shaler has advanced the theory that "many of the fissures were produced by the escape of gases, which broke forth with all the violence of volcanic eruptions, throwing out great quantities of sand and water." ^ Extrusion due to the closing of fissures is probably the true explana- tion of most of the more violent ejections, as it agrees with what would be expected in the case of earth fissures in a thin layer of aQuvium resting on a bed of quicksand opening and closing on the passage of large earth waves. It is, moreover, in harmony with the observations of reliable observers in several other more recent shocks. The explosive escape of gases generated by the decay of buried organic matter and disturbed by the earthquake vibrations is a competent and probable supplementary cause of the violent ejections. The quiet extrusion appears, on the other band, to be due to the local and temporary development of true artesian conditions, pre- sumably by the unequal settling of the deposits and the production of differential pressures. Although longer-lived than the extrusions described above, they generally last only a f ew minutes or hours. It is the outflow of waters by quiet extrusion that has carried the sand of the blows and sloughs. The absence of the blows on the higher lands, such as the prairie ridges, appears to be due in part to the greater thickness of the deposits above the quicksand beds and their greater resistance to fissuring, to their incoherent sandy character which tended to take up the earthquake movements by general readjustments among the grains rather than by fissuring, and to the f act that the head of the 1 BriDgier, L., Am. Jour. Sei., Ist ser., vol. 3, 1821, pp. 15-46. s Le Sieur, Qodfrey, ikrinted in Am. Geologist, vol. 30, 1902, pp. 79-80. s Biadbory, John, Eariy westem travels, Cleveland, 1904, voL 5, p. 209. * Shaler, N. S., Earthquakes of the westem United States: Atlantic Monthly, November, 1809, p. 555. tJNDEEMlNlNa. 87 water was generally insufficient to lift it to the higher surface, especially where numerous fissures in the adjacent bottoms afforded lower outlets. The absence of blows in the vicinity of the Missis- sippi River appears to be due to the f act that the waters and quick- sands found an escape by lateral flowage to the stream rather than by extrusion at the surface of the ground. UNDEBMININa. Favlt trenches, — The fault trenches, or canal-like depressions resulting from the settling of a fault block between two adjacent fissures, have been fully described elsewhere (p. 57). As there pointed out, it is probable, especially in the case of the larger trenches at right angles to the streams, that the dropping was permitted by an undermining due to flowage of the underlying quicksands toward the rivers, the Channels of which were in many places cut below the quicksand level. SaTid sloughs. — These broad shallow troughs, as has also been pointed out in another place (p. 83), are always marked by extensive accumulations of sand, and their formation, even when due to more general causes, has doubtless been assisted by local undermining due to the extrusion of the sand. Sinks, — Sinks are among the most conspicuous yet among the rarest of the material phenomena of the New Madrid earthquake, being reported principally in the vicinity of New Madrid. They may be described as circular depressions in the alluvium, originally from a few feet up to 15 yards or more in diameter and from 5 to 30 feet IQ depth. Lyell, in his account of his second visit to New Madrid,* says: Hearing that some of these cavities still existed near the town, I went to see one of them, three-quarters of a mile to the westward. There I found a nearly circular hol- low, 10 yards wide and 5 feet deep, with a smaller one near it, and I observed, scattered about over the surrounding level ground, fragments of black bituminous shale, with much white sand. Within a distance of a few hundred yards were five more of these sand-bursts, or sand blows, as they are sometimes termed here, and, rather more than a mile farther west, near the house of Mr. Savors, my guido pointed out to me what he called '^the sink hole where the negro was drowned.'' It is a striking object, intemipting the regularity of a flat piain, the sides very steep, and 28 feet deep from the top to the water's edge. The water now standing in the bottom is said to have been originally very deep, but has grown shallow by the washing in of sand, and the crumbling of the bank caused by the feet of cattle Coming to drink. I was assured that many wagon loads of matter were cast up out of this hoUow, and the quantity must have been considerable to account for the void; yet the pieces of lignite and the quantity of sand now heaped on the level piain near its borders would not suffice to fill one-tenth part of the cavity. Perhaps a part of the ejected substance may have been swallowed up again, and the rest may have been so mixed with water as to have spread freely like a fluid over the soil. 1 Lyell, Charles, A second visit to the United States of North America, London, 1849, vol. 2, pp. 232-233. 88 THE NEW MADRID EABTHQTJAKE. Similar sinks are also mentioned by Le Sieur.^ Still another description is by Bringier,' who says: The whole suiiace of the country lemained covered with holes, which, to compare small things with great, reeembled so nuiny craters of volcanoes, surrounded with a ring of carbonized wood and sand, which rose to the height of about 7 feet. I had occasion a few months after to sound the depth of Beveial of these holes, and found them not to exceed 20 feet; but I must remark the quicksand had waahed into them. The writer in going through other parts of the region saw nothing comparable to the sinks of New Madrid or those described by Bringier, although near Caruthersville the remains of small craterlets were observed in the open fields (PL IX, A) and small circular depres- sions 10 feet or so in diameter and 1 to 2 feet deep were noted at one point in the forest. As it was circular and not irregulär in shape and was not associated with any mound it can not be ref erred to the uprooting of a large tree. The feature to which the term sand blow is now applied is a low patch or ridge of sand, apparently extruded at the time of the earthquake, but not usually associated with any recognizable trace of the opening from which the sand came. The f act that carbonaceous matter, which must have been extruded from the ground, occurs about the sinks near New Madrid points to their being the orifices ''where the principal fountains of mud and water were thrown up" as stated by Bringier, the engineer, to Lyell.* As Lyell has pointed out, the sand heaped up about their borders would fill but a small part of the cavity and it is not impossible that the ejected material was so mixed with water as to spread freely like a fluid over the seil. That this was the case with much of the material extruded at the more recent Charleston earthquake was made appar- ent to the writer from his examination of the earthquake craterlets in 1905. There is, however, another explanation which the writer would suggest. This is as follows: At the start, the water in the underlying quicksand, which was under temporary pressure due to disturbance of the alluvium by the earth waves, rose through cracks or craterlets and spread sand and carbonaceous materials over the ground. Later, as the pressure was removed, the water subsided and the caving walls filled the cavity produced by the removal of the sand brought to the surface and obliterated the original fissures. Ordina- rily the story ended here; but, where the semifluid quicksand out- cropped in a near-by river bank, as is the case at New Madrid, there is a possibility of undermining through the flowage of the quicksand toward the stream. The resulting caving tended to be localized where the crust was already broken by fissures, although such localization is not necessary. The fact that the sinks are found only near the river and under conditions similar to those pos- tulated seems to poiat to the possibility of such an origin. Evidence 1 Le Bieor, Qodfrey, op. dt., p. 80. * Biincper, L., loc cit. * Lyell, Charles, op^ cit., p^ 23L AGITATION OF WATEE STJBFACES. 89 supporting this hypothesis is presented by similar sinks under like conditions in the alluvial deposits of Minnesota and other regions which we have no reason to think have been disturbed by earth- quakes HYDBOLOGIC PHENOMENA. Among the hydrologic phenomena associated with the New Madrid earthquakes may be mentioned the extrusion of water from the fissuresy the agitation of the water surf aces, the disturbances affect- ing nayigation, and the increase or decrease in the flow of Springs. The first of these is properly a phenomenon of extrusion and has been discussed at length under that head. The others, relating more to surface waters, are considered below. AGITATION OF WATER SUBFACES. Among the phenomena of the earthquake most appalling to the inhabitants of the New Madrid region was the behavior of the river at the time of the shock, and a number of graphic accounts of it have been handed down. As usual, one of the best descriptions is by the English naturahst Bradbury, who states that he was awak- ened at the time of the first shock — by so violent an agitation of the boat that it appeared in danger of upsetting. * * * I could distinctly see the river agitated as if by a storm. * * * Immediately the perpendicular surfaces [banks], both above and below us, began to fall into the river in such vast masses as nearly to sink our boat by the swell they occasioned. * * * The river was covered with foam and drift timber, and had risen considerably. * * * Two canoes floated down the river. * * * We considered this as a melancholy proof that some of the boats we passed on the preceding day had per- ished. Our conjectures were afterwards confirmed. Capt. Nicholas Roosevelt, already quoted, reported the river unusually swollen and turbid.* Just below New Madrid, according to Lloyd, a flatboat belonging to Richard Stump was swamped and six men drowned. At times the waters of the Mississippi were seen to rise up like a wall in the middle of the stream and suddenly roUing back would beat against either bank with ter- ri£c force. Boats of considerable size were of ten cast high and dry upon the shores of the river * * *. A man who was on the river in a boat at the time of one of the shocks declares that he saw the mighty Mississippi cut in twain, while the waters poured down a vast chasm into the bowels of the earth. A moment more and the chasm was filled, but the boat which contained this witness waa crushed in the tumultuous efforts of the flood to regain its former level.^ Hildreth, in relating the experiences of an eyewitness who was on the river 40 miles below New Madrid at the time of the earth- quake, says: * In the middle of the night there was a terrible shock and a jarring of the boats * * *. Directly a loud roaring and hissing was heard, like the escape of steam 1 Latrobe, C. J., Th6 rambler in North America, London, 1836, vol. 1, pp. 107-108. * Lloyd, J. T., Lloyd's steamboat directory, Cincinnatl, 1856, p. 825. s Hildreth, S. F., Original contributions to the American Pioneer, Cincinnati, 1844, pp. 34-35. 90 THE KEW MADRID EABTHQUAKE. from a boiler, acoompanied by the most violent agitation of the shores and ixe- mendouB boiling up of the waters of the Missiasippi in huge swells, roUing the waters below back on the descending stream, and tossing the boats about so violently tha( the men with diflSlculty could keep upon their feet ♦ ♦ ♦. The water of the river, which the day before was tolerably cleur, being rather low, changed to a reddish hue, and became thick with mud thrown up from its bottom; while the sur- iace laahed violently by the agitation of the earth beneath, was covered with foam, which, gathering into masses the size of a barrel, floated along on the trembling sur- JBce * * ♦. From the temporary check to the current, by the heaving up of the bottom, the sinking of the banks and sand bars into the bed of the stream, the river rose in a few minutes 5 or 6 feet, and, impatient of the restraint, again rushed for- ward with redoubled impetuosity, hurrying along the boats, now set loose by the horrornstricken beatmen, as in less danger on the water than at the shore where the banks threatened to destroy them by the falling earth or carry them down into the vortices of the sinking masses. Many boats were overwhelmed in this manner, and their crews perished with them. It required the utmost exertions of the men to keep the boat of which my Informant was the owner in the middle of the river as far from the shores, sand bars, and Islands as they could. At New Madrid several boats were carried by the reflux of the current into a small stream that puts into the river just above the town, and left on the ground by the retuming water a consid- erable distance from the Mississippi. A man who belonged to one of the Company boats was left f or several hours on the upright trunk of an old snag in the middle of the river, against which his boat was wrecked and sunk. It stood with the roots a few feet above the water, and to these he contrived to attach himself , while every fresh shock threw the agitated wavee against him, and kept gradually settling the tree deeper into the mud at the bottom, bringing him nearer and nearer to the deep muddy waters, which, to his terrified Imagination, seemed desirous of swallowing him up. While hanging here, calling with piteous shouts for aid, several boats passed by without being able to relieve him, until finally a skiff was well manned, rowed a short distance above him, and dropped downstream close to the snag, from which he tumbled into the boat as she floated by. Bryan^ says: At first the Mississippi seemed to recede from its banks, its waters gathered up like mountains, leaving boats high upon the sands. The waters then moved inward with a front wall 15 to 20 feet perpendicular and tore boats from their moorings and car- ried them up a creek closely packed for a quarter of a mile. The river feil as rapidly as it had risen and receded within its banks with such violence that it took with it a grove of cottonwood trees. A great many fish were left ux>on the banks. The river was literally covered with the wrecks of boats. An informant of Mitchill,^ who was on the Mississippi River, 87 miles below the Ohio, stated that the stream rose 6 feet from its former level and acquired three times its former velocity. William Shaler, also quoted by Mitchill,' says of the experiences of a friend: He immediately cut his cable and put oü into the middle of the river^ where he soon found the current changed, and the boat hurried up, for about the space of a minute, with the velocity of the swiftest horse; he was obliged to hold his band to his head to keep his hat on. On the current's running its natural course, which it did gradually, he continued to proceed down the river, and at about daylight he 1 Bryan, Eliza, quoted in Am. Geologist, vol. 30, 1902, pp. 77-78. * MitchiU, S. L., Trans. Lit. and Phüos. See. New York, yoI. 1, pp. 281-307. sidem,p.300. AGITATION OF WATBB SUBFACES. 91 came to a moet teirific fall, which, he thinks, was at least 6 feet perpendicülar, extending acroes the river, and about half a mile wide. The whirls and ripplings of this lapid were euch that his veesel was altogether unmanageable, and destruction seemed inevitable; some of the former he thinks, were at least 30 feet deep, and seemed to be formed by the water's being violenÜy sucked into some chasm in the river's bottom. He and his men were constantly employed in pumpiog and bailing, by which, and the aid of Providence, he says he got safe throngh. As soon as he was able to look round he observed whole forests on each bank fall prostrate, to use bis own comparison, like soldiers grounding their arms at the word of command. On his arrival at New Madrid he found that place a complete wreck, sunk about 12 feet below its level, and entirely deserted; its inhabitants with those of the adjacent coimtry, who had fled there for refuge, were encamped in its neighborhood; he represents their cries as truly distressing. A laige baige, loaded with 500 barreis of flour and other articles, was split from end to end and tumed upside down at the bank. Of nearly 30 loaded boats only this and one more escaped destruction; the water ran 12 feet perpen- dicular, and threw many of them a great many rods on shore; several lives were lost among the beatmen. Flint/ in describing the earthquake, says: A bursting of the earth just below the village of New Madrid, arrested this mighty stream in its course, and caused a reflux of its waves, by which in a little time a great number of boats were swept by the ascending current into the mouth of the bayou, carried out and left upon the dry earth, when the accumulating waters of the river had again cleared their current. Haywood ' says that on the first shock the — waters in the Mississippi near New Madrid rose in a few minutes 12 or 14 feet and then feil like atide. * * * Spoutsof water 3 or 4 inches in diameter sprang from the Miss- issippi to a great height. In some parte of the Mississippi the river was swallowed up for some minutes by the seeming descent of the water into some great opening of the earth at the bottom of the river. Boats with their crews were engulfed and never more heard of . Disregarding the element of fantasy m these descriptions, there is little room to doubt that the earthquake produced water waves of considerable size on the Mississippi. They doubtless started from the sideS; for the most part, and meeting in the middle produced the shaxp wall-like ''chop" described at this point. As usual, it was the retum wave which did the damage along the shore. That there were upheavals of the bottom is certain, as has been indicated in the dis- ciission of f aults and domes, and it is not unlikely that the water was thrown back, giving the appearance of a chasm. There is also no reason to doubt that jSssiures opened and closed beneath the water as they did on the land, giving rise to large waves by the ejection of water. That waves of great size moved upward against the current is certain, and that the movement of the water was retrograde for the moment, at least in shallow water, is probable. The rise in the river about which there is universal agreement was probably due in part to the waves moving upstream, and in part to a temporary ponding due to local uplifts of the river Channel. 1 Flint, Timothy, Reoollectioiis of the last ten years, Boston, 1826, p. 224. * Haywood, John, The natural and aboriginal history of Tennessee, Nashville, 1823, pp. 30-33 92 THE NEW MADBID EABTHQtTAKS. Not only were the waier surfaces agitated by the severe shocks, but in some localities, even to the east of the Mississippi, the pre- liminary tremors and minor vibrations produced noticeable effects on quiet surfaces. Thus Haywood says, ''the ponds of water, where there was no wind, had a troubled surf ace the whole day preceding any great shock." ^ The same feature was also observed by Jared Brooks, of Louisville, who, more than anyone eise, paid careful attention to the earthquake phenomena. He says ' he — stopped on the bank of a deep pond, the sur&u^e of which was a perfect mirror to appearance, oveihung by lofty trees; it instantly assumed the duU complexion and seemingly the roughness of a file; conveiging waves were soon laised by the quick motion of the shores, and, contending with each other, caused a curious commotion. The noise produced by the agitation of the trees resembled that of a ahower of small hail in the forest. Further away from the center of disturbance the action was less severe. Describing the trip of Capt. Nicholas Boosevelt, of New York, from Pittsburgh to New Orleans, Latrobe says* that while loading the boat from a vein of coal near Yellow Banks, Ind., the voyagers — were accosted in great alarm by the squatters of the neighborhood, who inquired if they had not heard stränge noises on the river and in the woods in the course of the preceding day and perceived the shores shake, insisting that they had repeatedly feit the earth tremble. It should be stated, however, that nothing is said of any agitation of the waters at this point. EFFECT ON NAVIGATION. The conditions of navigation on the Mississippi were much changed by the earthquake. The river was temporarily covered with wreck- age and d6bris, snags and sawyers multipUed, the banks caved, and islands disappeared. Caving of hariks. — Of the caving of the river banks the best accoimt is afforded by Bradbury,* who states: Immediately the perpendicular banks, both above and below us, b^an to fall into the river. I now saw clearly that our lives had been saved by our boat being moored to a sloping bank. Mr. Bridge, who was standing within the declivity of the bank, narrowly escaped being thrown into the river. as the sand continued to give way under his feet; the banks in several places feil within our view. Capt. Roosevelt, from the deck of the pioneer steamer New Orleans^ as he was passing down the Ohio along the Indiana shore — ever and anon heard a rushing sound and violent splash, and saw laige portions of the shore tearing away from the land and falling into the river. It waa, as my infonnant said, an awful day; so still that you could have heard a pin drop on the deck. 1 Haywood, John, op. dt., p. 124. * MoMortrie, H., Sketches of LouJsvüle and its environs, Loolsvffle, 1819, Appendix by Jared Brooks, p. 253. * Latrobe, G. J., The rambler in North America, London, 1836, voL 1, p. 107. * Biadbury, John, Early westem travels, Cleveland, 1904, voL 5, pp. 204-210. "'J&ri BPFEOT ON NATIQATION. 93 They had usually broi^ht to under the ahore, but everywhere thej saw tbe high ban^ disappeariug, OTerwhehning many a flatboat and raft, from which the owners had landed and made tbeir escape. Here they lay, keeping watch on deck during the long autumnal night, listening to the sound of the watera which roared and gui^led hor- ribly arotind them, and heaiing from time to time the nishii^ earth shde from the shore, and the commotion as the falling mass of earth and trees was swallowed up by the river. On reaching the Mississippi they found the Channel unrecognizable, everything heing changed by the action of the shock. Lloyd,' in describing the experiences of another Toyager, says that "during the TariouB shocks the banks of the Mis^ssippi caved in by whole acres at a time." Flint* Bays the graveyard at New Madrid was precipitated into the river. Ät New Madrid, according to Lyell,* the caving of the river was such at the time of the earthquake and in the following years that at the time of bis visit the river was fiowing over the site of the town. (See fig. 17.) Disappearance ofialands. — Many Islands in the Missis- sippi disappeared at the time of the earthquake. Hildreth says * "the sand bars and points of Islands gave way, swallowed Up in the tumultu- ous bosom of the river." La- trobe,' describing the trip of the first steamer on the river, says: "The pUot, alarmed and confused, affirmed that he was lost, as he found the Channel everywhere altered * * *. A large Island in mid-channel, which was selected by the pilot as the better alternative [for anchoring], was sought for in vain, having disappeared entirely." Dow * reports the washing away and disappearance of two Islands in the New Madrid area. Broadhead ' qUotes from the St. Louis namc 17.— Skelcli ihowiiig dunga fa chamwl of Ulssb slppf River at New Uadrld aCler tb« «arthqualas ot ISll-lZ. Lloyd, I, T., Lloyd's ateamboat dliecloiy, ClncIniiBtl, 18U, p. 321. FUnI, TEmothr, Recollsctloiu of tb« last len jeais, Boston, 183fl, pp. Z22-22S. Lyell, CliBriea, A second vislt to the United Btalea, vol. 2, pp. 228-229. HDdretb, B. P., Original oontribuckiia bi the American Pktneer, Clnclnnatl, ISM, pp. 34-3£. Latrobe, C. T-, Tbe nmblei in North America, London, 1B36, p. lOS. Dow, Lorenzo, quoted in Am. Qeologlst, voL 30, p. 77. :., Am. Qealoglat, voL 30, p. 83. 94 THE NEW MADBID EARTHQUAKE. Globe-Democrat of March, 1902, an artide from the papers of Aug. Warner on the disappearance of Island No. 94: ^' This island was in the Iower Afississippi, not f ar from Vicksburg. * * * jj^ the night the earthquake came, and next moming, when the accom- panying haziness disappeared, the island cotdd no longer be seen — it had been utterly destroyed." The disappearance of the islands, which were tisually only a f ew feet above the water, seems to be diie partly to washing and partly to the flowage of the loose, incoherent, water-sattirated sands of which they were composed. Snags and sawyers. — ^Besides the permanent changes in the banks, islands, and Channels, the ntimber of the snags and sawyers was increased or their position shif ted. The snags or planters, consisting of partly or wholly grounded wreckage — ^generally tree tnmks — and the sawyers, or those tree trtmks grounded at one end while the other rises and falls or sways from side to side in the current, were always dangerous to navigation and were particularly so in the few months following the earthquake. EQldreth ^ states that: ''Numer- ous boats were wrecked on the snags and old trees thrown up from the bottom of the Mississippi, where they had rested f or ages. * * * A man who belonged to one of the Company boats was left for several hours on an upright trank of an old snag in the middle of the river." The pilot of the New Orleans found, according to Latrobe,' the ''Channel everywhere altered, and where he had hitherto known deep water there lay numberless trees with their roots upward." Floating wrecka^ge. — ^Another conspicuoUs result of the earthquake was the increase in floating wreckage. Bradbury' describes the river as fuU of drift timber after the shock. He found the Devils Channel impassable ''from the trees and driftwood that had lodged during the night against the planters fixed in the bottom of the river." One correspondent of Mitchill* describes how "the trunks of trees, bedded in the bottom, suddenly rose in great numbers to the surface," during the shock, while Daniel Bedinger wrote to him "of the ele- vation of innumerable logs and trees from the bottom of the Missis- sippi." William Shaler, also quoted by Mitchill, mentions "a great multiplication of sawyers." ^ Combining the partly buried trunks set free by the agitation of the river bottom by the earthquake with the thousands of trees swept into the river by waves or by the caving of the banks, as described on page 96, the floating timber must have been enormous, adding greatly to the difficulties of navigation. 1 Hfldreth, & P., op. dt., p. 130. s Latrobe, C. J., The rambler in North America, London, 1836, vol. 1, p. 106. s Bradbnry, John, Early westem travels, CleTdand, 1904, vol. 5, pp. 20&-a07. * Mitchill, S. L., Trans. Lit. and Fhilos. Soc. New York, vol. 1, p. 293. » Idem, p. 302. THE NEW MADRID EABTHQUAKE. 95 CHANOES IN SPRINGS. Among the phenomena of the earthquake were certain changes in the character of particular Springs, often at distances remote from the Center of disturbance. Col. Samuel Hammond, in a letter to Mitchili/ states: In the county of Christian, Ky., a fine and fresh spring of water was observed to run very muddy for several hours. On examining it, after the feculence had settled, he foiind it to be strongly impregnated with sulphur, so much so that it was spoiled for domestic uses. Indeed, it had been converted to one of the strongeet brimstone Springs he ever met with. P. H. Cole, ako writing to Mitchill, says: In the month of September I visited a spring of about the distance of 14 miles from my residence. It was situated on the bsuik of a creek that issued forth streng sul- phurous water. The smell was evident to a considerable distance. It received its Bulphurons impregnation from a very heavy earthquake that occurred in January. Before that event it was a limestone water. On that occasion a new limestone spring broke out about 20 f eet above the original spring, and to this day the respective foun- tains pour forth their calcareous and sulphurous waters in distinct currents. Some Springs ceased to nm for some time, and others ran muddy several hours after the earth had been convulsed. In many places in west Tennessee, according to Haywood^' old sulphur Springs have commenced ninning again, which some years before were dried tip, while several new permanent Springs of sulphur water broke out. ACTION OF THE EABTHQTTAXE ON FOBESTS. The modification of the forests of the Mississippi Valley by the New Madrid earthquake was extensive and important^ and included the Splitting and overtuming of trees, their uprooting by landsHdes and caving banks, their destruction by rushing waters, the upUfting of wet-land species to high and dry positions, and the submergence of highland species.' SFLITTINa OF TBEES. The spUtting of trees is noted iu two of the contemporaiy descrip- tions, Bringier describing ''trees being blown up, cracked, and split and falling by thousands at a time/' and Dillard stating ''I have Seen oak trees which would be spUt in the center for 40 feet up the trunk, one part standing on one sidejof a fissure and the other part on the other, and trees are now standing which have been cleft in this manner." 1 Mitchill, S. L., op. cit, p. 292. * Haywood, John, The natural and aborigüial history of Tennessee, NashviUe, 1823, introdactory chapter. a FuUer, M. L., Earfhguakes and the forest: Forestry and Irrigation, voL 12, 1906, pp. 261'-287. 96 THE NEW MADBID EABTHQUAKE. DESTBUdnOK OF FOBESTS BT WATES AND OAYINO BANKS. The best account of the destniction by waves is that of Bryan, who mentions the faUing of trees and describes how an imusual wave of the river in receding ''took with it the grove of cottonwoods which hedged its borders. They were broken off with such regularity that in some instances persons who had not witnessed the fact could with difficnlty be persuaded that it was not the work of art." Lorenzo Dow adds that thousands of willows were swept off Uke pipestems about waist-high. Hildreth, in describing the convulsion on the Mississippi, says: ''The sand bars and points of Islands gave way * * * carrying down with them the cottonwood trees cracking and crashing, tossing their arms to and fro." Bradbury, who was on a boat on the Mississippi, says he could distinctly hear the crash of falling trees, while Devils Channel "appeared absolutely impassable from the quantity of trees and drif twood that had lodged there during the ni^t. * * * The banks in several places feil in, within our view, carrying with them innimierable trees, the crash of which, falling into the river, nuxed with the sound attending the shock * * * produced an idea that all nature was in a state of dissolution.^' A later shock likewise threw great numbers of trees into the river. The effect on the river is also brought out by Latrobe, who, in describ- ing Capt. Roosevelt's trip, says that after the shock the latter found the '* Channel everywhere altered, and where he had hitherto known deep water there lay numberless trees with their roots upward," while the ''trees were seen waving and nodding on the bank, without a wind." OVEBTHBOW OF FOBESTS BY LANDSLIDES. Besides the prostration of the forests by the caving of banks along the river, many trees were overthrown by landslides on the steep hill- sides, especially along the face of the Chickasaw Bluffs, which border the Mississippi lowlands in westem Tennessee. These bluffs are exceptionally steep, largely composed of clay, and when saturated with water, as they often are, present especially favorable conditions for landslides. Their fronts for miles are marked by landslip scars. A study of the age of the trees shows that the greater part of the upright growth on the disturbed surfaces is f airly uniform and a little less than 100 years of age, trees of greater age being in general tilted and partly overthrown. From this it seems clear that the main slides took place about 100 years ago, presumably at the time of the New Madrid shocks. The record of the action of the landsHdes is still preserved, as indicated, in the older trees as well as in the landslip features. The face of the bluff, which was already nearly as steep as the material could stand, seems to have hterally cnunbled under the ACTION OF THE EARTHQUAKE ON FORESTS. 97 action of the earthquake, great masses slipping downward, canying with them the immense primeval trees which covered the surface, mingling trees and earth in a confused jumble. Piate IV, -4, shows trees partly overthrown by the landsUdes at this time. In this case the original trees survived, gradually straightening until their upper parts again reached an upright position, while the lower parts remained inclined. In other instances, where the original tree was snapped off by the .hock, the main trunk has rotted away in the course of years, while a new tree, developed from one of the old Umbs or shoots, has taken its place. GENERAL PROSTRATION OF FORESTS BY VIBRATIONS. Besides the destniction of the forests by water, landsHdes, etc., there seems to have been a general prostration in certain iocalities due to the vibrations alone. For instance, WiUiam Shaler, in describ- ing a friend's experiences, says he saw ^'whole forests on each bank faÜ prostrate * * * ]ike soldiers grounding their arms at the Word of command." Brmgier, in describing bis experiences to LyeU,' said that in some of the severest shocks, as the waves advanced, he saw the trees bend down and often, the instant afterwards, when in the act of recovering their position, meet the boughs of other trees similarly inchned, so as to become interlocked, being. prevented from righting themselves again. In a published account the same observer^ teils of ''the horrible disorder of the trees, which everywhere encountered each other, being blown, cracking and Splitting, and falling by thousands at a time." Very similar conditions in the vicinity of Little Prairie (Caruthers- ville) are reported by James.' ''The forest adjoining the settlement at Little Prairie, below New Madrid, presents a singular scene of con- fusion, the trees standing inclined in every direction and many having their tninks and branches broken.'' That such prostration occurs was substantiated by the writer's observations at Keelfoot Lake, over extensive areas of which the tniuks are prostrate, although elsewhere standing as upright and stiff as bef ore the shock. As there was no water in the region at the time, and as the land was nearly flat, only the vibrations or accompanying earth waves remain to account for their destniction. In places dozens, if not hundreds, of prostrate trunks can be counted. For comparison, reference should be made to Plate VII, -4, which shows the upright stumps normally found. 1 Lyell, Charles, A secoud Visit to tbe United States, vol. 2, p. 231. • Bringier, L., Am. Jour. Sei., Ist ser., vol. 3, 1821, pp. 15-46. * James, Edwin, Account of an expedition from Pittsbm^h to the Rocky Mountains, Philadelphia, 1823, vol. 2, p. 326. 98 THE NEW MADRID EABTHQUAKE. DEAD TBEES. A phenomenon of special interest, but somewhat puzzling, is described by Lyell.^ Skirting the Bayou St. John, he observed — a great many ftJlen treea and oüiers dead and lifeless but standing erect. * * * He took me to part of the forest, on the borders of what is called the " sunk country / ' where all the trees prior to 1811, although standing erect and entire, are dead and leaf- lesB. They are chiefly oaks and walnuts, with trunks 3 to 4 feet in diameter, and many of them 200 years old. They are supposed to have been killed by the loosening of the roots during the re])eated undulations which passed through the soil for three months in succession. UPLIFTED TREES. The uplifted trees, which are occasionally seen in the Reelfoot Lake district, are of special interest because of their rarity, the phe- nomena of subsidence being much more common. In Plate VII, B, however, is represented an old cypress with a füll development of knees, denoting f ormation in water, the tree being now on high and dry grassy ground, entirely above the reach of the water, except pos- sibly in the time of the highest floods. SUBMEROED FORESTS. The phenomena of submerged f orests are among the most striking of the earthquake features, and were the subject of some of the most graphic descriptions of the early writers. Even to-day they con- stitute one of the most conspicuous classes of evidence of the shocks. The best developments are found in the sunk lands of the Vamey Kiver in Missouri Big Lake and Lake St. Francis in Arkansas, and in Reelfoot Lake in Tennessee. Many other localities exhibit similar features on a smaller scale. In f act, probable examples may be found in nearly all of the bayous and sloughs connected with the St. Fran- cis River, and their occurrence can not be better seen than by refer- ence to the accompanying map (PI. I), which shows the swamps and submerged lands as determined by the Mississippi River Com- mission and verified by the field work for the present report. Refer- ence should be made to the section on *'Sunk lands'' (p. 64) for detailed descriptions of the phenomena. AREA OF DESTRUCTION. ' Over how great an area the f orests were destroyed it is dijQäcult tO| say at the present time. Reelfoot Lake alone probably originally covered 75 Square miles of forest, and the swamps formed at that time west of the Mississippi probably covered 125 Square mileq more. These two localities alone would comprise more than 125,000 Lyell, Charles, A seoond visit to the United States, vol. 2, pp. 234-236. xmoT Dir AxnncuiÄ siruotuses. 99 acree that wero destroyed. The amount of timber lost by the caving of river banks and by the OYerwhehniiig of the Islands would probaUy bring the totat to 160,000 acres. To this amount still further additions muat be nuule of the areas in which the timber was OTerthrown by landaUdes or other related causes. No estimate of tbe latter can now be made but it was undoubtedly considerable. That the total destruction was suffident to give earthquakes a place among the enemiiw of the forest can not be disputed.^ BFVBCfT OH ABTJJfiOZAIi 8TBTT0TXJBES. NEW ICADSm REGION. Relatively few details have been preserved in regard to the action of the earthquake upon buildings in the New Madrid region, the natural phenomena evidently making a far greater impression on Ute minds of the observers than the action on artificial structures. It is known, howeyer, that at the first shock chimneys and other structures of like nature and all loose objects were thrown to the ground, while the damage even to the low cabins was such that it 1>ecame dangerous to remain in them, and the inmates were obliged to rush out of doors. The larger and more pretentious houses f ared quite as badly. Many of the poorly built cabins and houses were destroyed and many fences were thrown down even by the first shock; while certain of the later shocks were even more destructive because of the weakening that all structures had suffered from the preceding shocks. Flint* states that at New Madrid most of the houses were thrown down in an hour, the crumbling being particularly severe during the passage of the large earth waves. As the earthquakes constantly recurred people no longer dared to dwell in houses or to use chim- neys, but passed that winter and those succeeding in light bark huts, tents, and other temporary shelters, using camp fires or low ovens to cook their food. The Little Prairie settlement (now Caruthers- ville) was broken up and the Great Prairie settlement was practically demolished. According to Stoddard ^ the town of New Madrid was ^'originally so laid out as to extend, as the French express it, 40 acres in length along the river; the back part was contracted to 20 acres on account of some swamps, and the depth was 16 acres. It contained 10 streets »Tunning parallel with the river, and 18 others crossing at right angles. The former were 60 feet and the iatter 45 feet in breadth. Six Squares were also laid out and reserved for the use of the town, »each of which contained 2 acres. * * * j^ street of 120 feet in 1 Füller, M. L., Earthquakes and the forest: Forestry and Irrigation, vol. 12, 1906, pp. 261-267. « Flint, Timothy, Recollections of the last ten years, Boston, 1826, p. 222. 3 Quoted by Beck, L. C, Gazetteer of the States of Illinois and Missouri, Albany, 1823, pp. 299-300. 100 THE ITEW MADRID ELABTHQUAKE. breadth was likewise resenred on the bank of the river." In 1799 the town had upward of 800 inhabitants, but at one time after the earthquake only two families remained, the remainder having fied. The public works and several streets were carried away by the cav- ing of the river banks and ' 'houses, gardens, and fields were swallowed up." The formation of ponds and the extrusion of sand and water f rom fissures added to the ruin. That the destruction was no greater in view of the severity of the shocks was due to the character of the buildings. Speaking of the destruction Flint says: ^ I infer that the ahock of these earthquakes in the immediate vicinity of the center of their force must have equaled in their temble heavings of the earth anything of the kind that has been recorded. I do not believe that the public have ever yet had any adequate idea of the violence of the concusEdons. We are accustomed to measure this by the buildings overtumed and the mortallty that results. Here the country wafl thinly settled. The houses, fortunately, were fndl and of logs, the most difficult to overtum that could be constructed. DISTANT LOOAL.ITIES. The destruction by the earthquake^ though much less than at the center of the disturbance near New Madrid, extended many hun- dred miles. Some of the more definite reports of injuries, mainly based on reports of Mitchill,' are given below. St Louis, Mo. — Several chimneys were overthrown and anumbci of stone houses split. Hercylaneum, Mo, — Brick and stone chimneys were injured and some broken oflf and thrown down by the strenger shocks. Oape OirardeaUj Mo, — Several houses were thrown down.^ NatcJiez, Miss, — ^The plastering in some houses was cracked. CarOvage, Tenn, — ^Bricks were thrown from chimneys and several were broken and overthrown; the brick courthouse was. cracked to the foundation and otherwise injured. ClarksviMey Tenn, — ^Many chimneys were injured. Henderson County, Ky, — Nearly every brick or stone chimney in the county was overthrown. Red Bank (150 miles heUm LouisviUe), — Several chimneys were thrown down and others damaged so as to be dangerous. LouisvMe. — ^The damage by the first shock was considerable, gable ends, parapets, and chimneys of many houses were thrown down. Similar destruction was produced by each of the three shocks of ^'great severity" recorded by Jared Brooks for the week ending December 22, by one in the week ending January 26, one in that end- ing February 2, and three in that ending February 9.* 1 Flint, Timothy, loo. cit. s MitchJU, S. L., Trans. Lit. and Phik». Soc. New York, vol. 1, pp. 281-307. s James, Edwin, Account of an expedition from Pittsburgh to the Rocky Mountains, Philadelphia, 1823, vol. 2, p. 326. * Casseday, Ben, History of LouisviUe, LouisviUe, 1852, pp. 121-126. NOISES. 101 Oindnnati, Ohio. — ^The tops of several chimneys were thrown oiF in town at the time of the first shock, while similar results probably attended the equally violent shocks of January 23 and 27. On February 7 more chimneys were thrown down and wide fissures made in brick walls.* South Carolina, — Cracked and started chimneys were reported at Laurens and Newberry. Georgia. — ^Bricks are reported to have been thrown from cliimneys. OTHEB PHYSICAL PHENOMENA. NOISES. The noises accompanying the more important shocks are among the most noticeable features of the earthquake. Bradbury teils of being *'awakened by a tremendous noise" at the time of the first shock and noted that ''the sound, which was heard at the time of every shock, always preceded it at least a second and uniformly came from the same point and went off in the opposite direction.'' Audubon, speaking of one of the more severe shocks, describes the sound as like ''the distant rumbling of a violent tornado." Bringier does not mention the subtertanean noise, but describes the "roaring and whistling produced by the impetuosity of the air escaping from its confinemenf in the alluvial materials. A similar roaring and hissing, like the escape of steam from a boiler, is also described by Hildreth. Eliza Bryan describes the sounds of the shock itself as "an awful noise resembling loud and distant thunder but more hoarse and vibrating. '' Flint compares the sound of the ordinary shocks to rumbling like distant thunder, but mentions the fact that the vertical shocks were accompanied by "explosions and a terrible mixture of noises. '' Linn describes the phenomena as beginning with "distant rumbling sounds, succeeded by discharges, as if a thousand pieces of artillery were suddenly exploded," and notes the hissing sounds accompanying the extrusion of the water. In Tennessee, as described by Haywood,^ "in the time of the earthquake a murmuring noise, like that of fire disturbed by the blowing of a bellows, issued from the pores of the eartli. A distant rumbling was heard almost without intermission and sometimes seemed to be in the air/' Explosions like the discharge of a cannon a few miles distant were also recorded, and " when the shocks came on the stones on the surface of the earth were agitated by a tremulous motion like eggs in a frying pan, altogether made a noise similar to that of the wheels of a wagon in a pebbly road." 1 Drake, Daniel, Natural and Statistical view or picture of Cincinnati, Cincinnati, 1815, pp. 233-235. - Hay wood, John, Natural and aboriginal history of Tennessee, Nashville, 1823, opening chapter. 102 THE NEW MADRID EABTHQUAKE. In the more distant localities we must again rely on the compila- tion of Mitchill.^ At Herculaneum, Mo., the noise was described as a roaring or rumbling resembling a blaze of fire acted upon by wind; at St. Louis before the shocks ''sounds were heard like wind rushing through the trees but not resembling thunder," while more con- siderable noises accompanied the shocks. At Louisville the noisc was likened to a carriage passing through the street; at Washington, D. C, the sound was very distinguishable, appearing to pass from Southwest to northeast; at Richmond similar noises were heard, but they appeared to come from the east; at Charleston there was '^a rumbling Uke distant thunder which increased in violence of sound just before the shock was feit;" while at Savannah the noise was compared to a rattling noise hke ^'that of a carriage passing over a paved road." From the above it would appear that earth noises were heard at most points where the earthquake was feit. In the region of marked disturbance there were the additional noises made by escaping air, water, crashing trees, and caving river banks. According to the best information the sound in the Mississippi Valley was a some- what duU roar, rather than the rumbling sound of thunder with which it was compared at certain of the more remote localities. In reality, as has been stated to the writer in regard to the recent Jamaica earthquake, although suggesting many of the common noises, it was essentially unlike anything ever heard by the observers before. ORIGIN AND CAUSE OF THE NEW MADRID EARTHQUAKES . POPITI^AB BELIEFS. Of the various explanations oflfered by those who speculated on the causes of the shock the volcanic theory was, as usual, by far the most populär. The general nature of volcanic manifestations was more or less understood by everybody, while relatively few had heard of the slower processes of warping and f olding and the occurrence of f aulting and even those who had seldom appreciated their importance. It is but natural, therefore, that they should turn for an explanation to the known cause which seemed to explain the phenomena best. It is probably to this general beUef in the volcanic origin of the shock that the stories of the emission of sparks and similar phenomena are to be attributed. Even at the time, however, there were some who did not accept the volcanic theory. James,^ who accompanied Loi^g's expedition to the Rocky Mountains, says: 1 Mitchili, S. L., loc. dt. 2 James, Edwin, Account of an expedition from Pittsburgh to the Rocky Mountains, Philadelphia, 1823, vol. 2, pp. 325-326. CAUSE OF THE EARTHQUAKES. 103 It has been repeatedly afiserted that volcanic appearances exist in the mountaiiioua country between Cape Girardeau and the Hot Springs of the Washita, particularly at the latter place; but our Observation has not tended to confirm these acconnts, and Hunter and Dunbar, who spent some time at the Hot Springs, confidently deny the existence of any such appearances in that quarter. Keports have been often circu- lated, principally on the authority of hunteis, of explosions, subterranean fires, blow- ings and bellowings of the mountains, and many other singular phenomena said to exist on the Little Missouri of the Washita and in other parts of the region of the Hot Springs, but it is eaay to see that the combustion of a coal bed or something of equal ingignificsCnce may have afforded all the foundation on which these reports ever rested. Speaking of the New Madrid shock Bringier says: ^ Several authors have asserted that earthquakes proceed from volcanic causes, but although this may be often true the earthquake alluded to here must have had another cause. Time perhaps will give us some better ideas as to the origin of these extra- ordinary phenomena. It is probable that they are produced in different instancea by different causes and that electricity is one of them; the shocks of the earthquake of Louisiana in 1812 produced emotions and sensations much resembhng those of a streng galvanic battery. It will perhaps be pertinent to observe that this earthquake took place after a long succession of very heavy rains, such as had never been seen before in that country. Nuttall,* on the other hand, refers the earthquake to '^the decom- position of beds of lignite or wood coal saturated near the level of the river and fUled with pyrite.'' In the light of our present knowledge of earthquake causes in gen- eral and of the nonvolcanic nature of any of the adjacent regions the theory of volcanic origin can be dismissed without further con- sideration. Electricity can likewise be mied out as a cause^ although certain electrical and magnetic phenomena appear to be associated with most great earthquakes. Aside from the absence of records, geologic or other, of any extensive expulsion of gas we have the testimony of all eyewitnesses to the effect that the expul- sion of gases and the associated water was the result and not the cause of the disturbance. EVIDENCE OF OBIGIN. One of the best evidences of the origin of the New Madrid earth- quake is the nature of the vibrations. If they had resulted from a disturbance within the alluvial deposits themselves, it could have been only from such causes as the expulsion of water and gases, the caving of banks, the slipping of landslides, etc. In regard to thefirst, it need only be repeated that there is no evidence that any extrusions of water or gas, except those arising in consequence of disruptions brought about by (not causing) the shocks. Of the caving of banks it may be said that had the shores of the Mississippi caved from source 1 Bringier, L., Am. Jour. Sei., Ist ser., vol. 3, 1821, p. 20. s Nuttall, Thomas, quoted by James, loc. cit. 104 THE NEW MADBID EABTHQUAKE. to mouth the vibrations would have been feit only a few hundred feet. Similarly in regard to the landslides. Although the f aces of the Chick- asaw Bluffs were affected f or miles, it was by slipping and not by direct fall, and the jar, if any was produced at all, could have been feit only a Short distance. The vertical yibrations accompanying several of the shocks were of a natnre differing from any which we can conceive as developing in the unconsolidated deposits themselves. The best evidence of origin is that afforded by the distance to which the yibrations were feit. It does not seem possible to conceive of a shock originating in soft embayment deposits being transmitted to the hard rocks and thence across the Appalachians to the Atlantic coast on the east and across the central coal basin to Chicago, Detroit, and Canada. The fact that the shocks were strongly feit at these localities seems to point conclusively to a deep-seated origin in rigid rocks. A f aulting in the hard Paleozoic rocks seems, therefore, to be the only probable explanation. LOCATION OF CENTBUX. Faults, some apparently of relatively recent date, are common in the Ozark Mountains which border the New Madrid area on the north- west and in the southem Illinois and northwestem Kentucky ^region on the northeast, and the thought naturally suggests itself that some movement in these regions, either a new or the readjustment of an old fault, may have been the cause of the shock. This view was advo- cated by Prof. Shepard,^ and was accepted by the writer in an earlier paper.* Further study in the field, however, has brought out the fact that the region of marked disturbance is confined to a certain definite area extending from southeastem Missouri to the latitude of Memphis and from St. Francis River on the west to Chickasaw Bluffs on the east, and nowhere reaching within many miles of the edge of the embayment area of soft deposits, as would have been the case if the shock had originated in the hard rocks outside this area. The gen- eral trend and shape of the area, taken in connection with the direc- tion of the earth waves, points to a centrum of the original shock along a line having the position indicated on Platel — ^that is, anorth- east-southwest line extending from a point west of New Madrid to a point a few miles north of Parkin, Ark. The centrum of the heavier subsequent shocks seems also to have been along essentially the same Une. The location of the centrum of some of the later and lighter shocks may have been elsewhere. The motion of one of the shocks at Detroit, for instance, is described as '^a poimding up and down instead of oscillating," ^ which could hardly have been the case if 1 Sliepard, E. M., The New Madrid eartliquake: Jour. Geology, vol. 13, 1905, p. 61. * Füller, M. L., Science, new ser., vol. 21, 1905, pp. 349-350. > Mitchill, S. L., Trans. Lit. and Philos. Soc. New York, vol. 1, p. 297. SIGNIFICANCE OF DOMINO. 105 the shock originated at so great a distance as New Madrid. Similar vertical shocks at Louisville are recorded by Brooks.^ Drake also notes' that diiring the second year of the earthquake shocks the Center of disturbance ''seems to have ascended the Mississippi to the Ohio and then advanced up that river about 100 miles to the United States Saline^ at which place shocks have been feit almost every day for neariy two years." It is also important in this connection to note that many of the later shocks, while not materially more severe than the lirst in the New Madrid region, were notably strenger in the more remote regions and were recorded at points in the east, from which no report of the first shock was made. It seems not improbable, then, that there was more than one Center of disturbance northeast of New Madrid in the years following the first shock, probably resulting from disturbance of the local equilibrium by the vibrations originating in the New Madrid area. ULTQCATE CAUSE. As shown elsewhere (pp. 109-1 10), the New Madrid earthquake was but one of a series that is still imfinished, indicating that in all proba- bility it resulted from causes th at are still active. Inasmuch as the cen- ter of activity of the primary shocks is within the embayment area and well removed from the surroimding areas of consolidated rocks, it seems clear that the ultimate cause lies in forces operating beneath the embayment deposits. The action may be associated either with the processes of folding or warping or incident to a depression and deepening of the basin. In this connection the phenomena of uplift and doming are of the greatest significance and may be reviewed and summarized to advantage. SIGNIFICAKCE OF DOMINa. It is of interest to note that a line drawn through the Tiptonville, Blytheville, and Little River domes agrees in trend with each of the major features of the earthquake region (PI. I), paralleling the line of the Chickasaw Bluffs and the direction of the Missisippi River on the east and the lines of the prairie ridges, sunk lands, and Crowley Ridge on the west. If the uplift were a few feet higher, a long ridge analogous to the prairie ridges and simulating on a small scale the larger Crowley Ridge would result, suggesting the possibility of a com- mon relationship due either to structure or to erosion. The Little Prairie Ridge, extending from New Madrid northward for neariy 25 miles, is a direct continuation topographically of the Tiptonville dome, and inasmuch as the river banks near New Madrid 1 Brooks, Jared, in History of Louisville by Ben Casseday, 1852, pp. 121-126. 2 Drake, Daniel, Natural and Statistical view or pictuie of Cincinoatit Ciodmiati, 1815, p. 237. 106 THE NEW MADRID EABTHQUAKE. show a doming of the alluvial depoeits parallel to the trend of the prairie; thero are good reasons f or thinking that the latter may repre- sent the structural as well as the topographic continuation of the dorne mentioned. If this is so, it foUows that the siinilar prairie extending from Big Lake to beyond Maiden may likewise be of structural origin. On this assumption there would be, beginning on the east, a structural trough occupied by Mississippi River and Keel- foot Lake; a structural ridge marked by the Tiptonville, Blytheville, and Little River domes, a second structural trough occupied by Little River, Big Lake, and Lake Nicormy, a second ridge represented by the Big Lake-Kennett-Malden Prairie, and a third trough occupied by St. Francis River and Lake. Still west of this is the topographic elevation of Crowley Ridge and the basin of Black River. In other words, there is a parallel series of four troughs and three ridges, all of which but one ridge and one trough are believed to be primarily structural in nature. The question arises, may not the two latter be also of structural origin, even if their magnitude is considerably greater? If we disregard the loess of Crowley Ridge a considerable part of the height is taken away, and we have a ridge only about 100 f eet high at the most and differing in no way except in size and the extent of its erosion from the lower prairie ridges, and it does not require a great effort to imagine a similar origin for the major ridge. Crowley Ridge has been greatly altered by erosion. Black and other rivers on the west have deepened the associated basins, thus accen- tuating the elevation of the ridge, and a similar work has been per- formed by St. Francis, Little, and Mississippi rivers on the east. The writer does not wish at this time to urge a structural origin for Crowley Ridge, but simply desires to call attention to the occurrence of smaller parallel lines of uplift and depression due to relatively recent warping and to suggest that an earlier and stronger warping of the same nature and in the same direction might have produced a simüar uplift along the Crowley Ridge axis, which determined the direction of drainage and the position of the resulting erosion remnant. If this is true, we have the beginning in Mississippi Valley at this point of what may ultimately develop into a syndinorium of great geologic importance. CAUSE OF XTPLIFT AND DOMING. The uplifts may be conceived as (1) representing the surfaoe indi- cations of disturbances originating in the underlying hard rocks; (2) due to unequal subsidence resulting from undermining; (3) due to horizontal flow of certain of the underlying deposits; and (4) having resulted from lateral movements tending to produce compression. That the uplift and doming is due to local arching of the bottom of the basin in which the embayment deposits occur is believed to be CAUSE OF UPUFT AND DOMINO. 107 imlikely, as such movements if communicated through a thousand feet or more of Sediments as they needs must be, would be likely to be of more general extent. Moreover, such arching of the hard rocks is a very slow process, while there is reason to believe that the doming took place with considerable suddemiess, possibly to a considerable extent during a single period of earthquake disturbance. Undermining, wMch may be conceived as resulting f rom the extru- sion of sand f rom the fissures or by flowage toward the rivers, seems to be even more improbable. It is true, as described elsewhere, that immense amounts of sand were extruded at the time of the shocks: enough, in f act, to cover nearly the entire surf ace over hundreds of Square miles. The layer, however, is usually only 3 to 6 inches deep, and is entirely insufficient to account f or the subsidence which in many places amounts to several feet. Locally there was doubtless considerable flowage of the substratum of quicksand toward the streams, but there was no general choking of the rivers from the cause, as would have been the case if amounts sufficient to account f or the subsidence had entered them. If there is no adequate basis for the theory of undermining, the latter can not be considered as a cause of diflferential movement such as might have given rise to domes by the subsidence of the.surrounding material. Horizontal flowage within the substratum of quicksand undoubtedly took place locaUy, as where there were general movements of the river borders toward the Channels, and it is conceivable that movements on a much larger scale may have occurred. Under this hypothesis the domes are to be considered as correlatives of the sunken areas, both being due to movements within the semifluid quicksand. The size of the domes and troughs seems to demand the diff erential move- ment of material equivalent to a body of sand 10 feet thick (half of the amount of the diff erence of surf ace elevation) for distances amount- ing to many mUes, the distances of the axes of the domes from the adjacent trough axes being from 5 to 10 miles or more. Such a trans- fer of material is believed to be improbable, although perhaps not impossible. The movement of the quicksand is thought by the writer to have been limited, in most places at least, to half the wave length of the surface undulations accompanying the shocks. In other words, the movements of the quicksand during the shocks are thought to have been sinülar to the water movements in ordinary waves, the sand particles moving in orbits of similar form without much lateral displacement. If the doming had been produced by lateral thrust communicated from the hard-rock borders of the basin, we should expect some signs of warping in the broad Tertiary plateau, especially on the east side of the Mississippi, but nothing of the sort is known. If, however, the cause originated immediately under the Mississippi bottoms, the 108 THE NEW MADBID EABTHQUAKE. alluvial deposits might be considerably affected without there being any appreciable effect on the uplands. On this point we have the evidence of the earthquake shock, which points to a seat of disturb- ance along a northeast-southwest line a few miles to the west of the Mississippi. If the fault giving rise to the earthquake was simply the outward expression of a movement of subsidence, as it may well have been, we may conceive of conditions similar to those shown in figure 18, in which D represents a depression formed either through the influence of the faulting or giving rise to it. The result of such a depression would be a settling of the overlymg deposits accompa- nied by a general lowering of the surf ace about it and a lateral move- ment in the direction of the arrows toward the center of disturbance, the result of which movement would be the formation of corrugations such as are represented by the domes and aUied uplifts. As suggested on page 106, even Crowley Kidge may have originated from sindlar FiQXTRE 18.— Diagram illustrating possible cause of doming in the New Madrid area. causes, although its present relief is due largely to erosion. This hypothesis explains the position of the domes and their absence from other areas on the east and west, the anticlinal nature of the disturb- ance, and accounts for the general preponderance of subsidence over uplift. On the whole it seems to be the most probable of the hypothe- ses suggested to account for the phenomena of warping and for the earthquakes, which appear to be an incident rather than a general cause of the changes going on. Although the general movement along the Mississippi is downward, at least differentially, the movement is not continuous. It is prob- able that as late as the deposition of the Lafayette gravels, in late Phocene or early Quatemary times, the land stood relatively lower than at present, the deposits named being laid down across the embayment area at an elevation of many feet above the present flood-plain. The distribution of loess affords some reason to think that similar conditions existed nearly down to the time of the Wisconsin glacial stage. If so, there lias been removed in relatively POSSIBILITIES OF FUTTJBB DISTtTEBANCES. 109 recent times a very considerable load from the area^ as a result of which the crust would tend to rise until an equilibrium was reached. It is even possible that the shocks of 1811-12 were incidents of an uplift rather than of depression. CONTEMPORANEOUS DISTURBANCES. The years 1811, 1812, and 1813 were characterized by unusual seismic activity in many other parts of the Western Hemisphere, as well as at New Madrid. On March 26, 1812, a terrific earthquake occurred at Caracas, throwing down churches and other buildings and destroying 10,000 persons. A little over a week later another heavy shock produced geologic effects of considerable magnitude, over 300 feet of the top of one of the mountains being lost by a land- sHde. Other shocks were feit at New Grenada and other of the West India islands, while a very important earthquake occurred in Cali- fornia at about the same time. Yolcanic action was similarly active. On April 27 the eruption of St. Vincent Soufriere began. During the year preceding more than 200 earthquake shocks presumably of vol- canic origin had been feit. Farther away, the island of Sabrina in the Azores was built up to a height of over 300 feet above the sea by a submarine eruption. A comparison of the dates of the great South American earth- quakes with those in Mississippi Valley does not suggest any direct connection. Neither March 26 nor April 5, the dates on which the principal shocks occurred, was characterized by any pronounced dis- turbance at New Madrid; in fact, both days were unusuaUy quiet. There is a tendency after any great disturbance to produce read- justments tp new conditions in surrounding areas, and a series of shocks may be expected to foUow in other parts of the same geologic province. There seems, however, to be no possible connection between the New Madrid and Caracas regions, the geologic provinces being entirely unrelated. That the disturbances may each be a sur- face expression of a single deep-seated cause, such as a general crustal or subcrustal readjustment, nevertheless, is not altogether improbable. POSSIBILITIES OF FUTURE DISTURBANCES. In another part of this report a hst of the earUer shocks experienced in the New Madrid region is given, and evidences are presented of the existence of earthquake cracks antedating those of 1811 (pp. 11 to 13). In connection therewith should be mentioned the corrobo- rative evidence afforded by Indian tradition, which teils of a great earthquake that had previously devastated the same region.* 1 Lyell, Charles, A seoond visit to the United States, vol. 2, p. 238. 110 We hav line of mij that the ci the ISll e aüces are i gies indic since in al sions, whe: has been f a hundred fulfilled. take place will occur Any sev disturbanc tucky; Car and Jonesi which are would be f lence of b] The larger Memphis, ' whole reg on a loess Mississippi saturated gave rise owii^ to t accentuatf Mississippi eartbquak the New 1 almost cei along the '' are less lii Madrid ar Judging fr probably ■ might resi Mt the year 1852," hy Ben Caseeday, Louieville, 1652, pp. 121-126. Deecribes the eartbquake phenomens »X I^uieville and gives a liet and clasei- fication of the ehocks. . In " Sketches ot Louisville and ite environe," by H. McMurtrie, Louisville, 1819, appendix, pp. 233-255. Gives a complete liat of obaerved ahockH, pendulum obaervationa, weather conditioiu, etc., at Louisville from Becember 16, 1811, to May, 1812. Bryan, Eliza. The New Madrid earthquake, abstract by G. C. Broadhead: Am. Geott^jst, vol. 30, pp. 77-78. See also Dow, Lorenso. Caxr, Ijucieii. American commonwealths; Miasouri, a bone of contention, Boeton, 1888, pp. 108-112. Givee an account ot the New Madrid earthquake compiled trom old narra- tivee, without credit, and describee frauds following the action of Congress in granling new lands to Bufferers. Casaeday, Ben. The history of Louisville fiom its earliest eettlement tili the year 1852, Louisville, 1852, pp. 121-126. See Jared Brooka. Cramer's Navigator, Pittsburgh, 1821, p. 243. Notes that the river bank for 15 miles above New Madrid sank 8 feet. DOlard, A. N. In ''The MiBsiaaippi Valley, ita phy Biograph y," by J. W. Foster, Chic^o, 1869, pp. 19-22. Reiatea a number of incidenU of the earthquake and deacribea the phenomeoa of the Bubmerged lands. Dow, Ziorenzo. "Lorenzo Dow'b works" {not seen), Cincinnati, 1850. Printa a letter from Eliza Bryan giving detaiied description o£ the earthquake 112 THE NEW MADRID EARTHQUAKE. Drake, DanieL Natural and Statistical view or picture of Oincinnati, Cincinnati, 1816, pp. 233-244. Includes a chronological list of the shockB, a Classification by intensities, a genend sununary of the history of the earthquake, an account of previoua dis- turbances, and a discussion of the contemporary electrical and physical phe- nomena and the State of the atmosphere at the time of the principal ehocks. Flinty Tiniothy. Recollections of the last ten yeais * * * in the Valley of the Mississippi, Boston, 1826, pp. 222-228. Gives a graphic story of the New Madrid shocks and remarks on the conditions existing on his visit seven years later. . A Condensed geography and history of the westem States in the Mississippi Valley, Cincinnati, 1828, 2 vols. Gives same account as in '* Recollections of the last ten years.'' . Extract from travels of Mr. Flint: Am. Jour. Sei., Ist ser., vol. 13, 1829, pp. 366-368. An abstract of the account in ''Recollectiona of the last ten yeara." The history and geography of the Mississippi Valley, Cincinnati, 1832, 2d edition, vol. 1. See "Condensed geography," by Flint. FoBter, J. W. The Mississippi Valley, its physical geography, Chicago, 1869, pp. 19-25. Quotes the accounts of A. N. Dillard and Timothy Flint at length, appending a few original paragraphs. (See Dillard, A. N., and Flint, Timothy.) Füller, Myron L. The New Madrid earthquake, by Edward M. Shepard : Am. Geolo- gist, vol. 35, 1905, pp. 180-181. Review of a paper in Jour. Geology, vol. 13, 1905, pp. 45-62. ■ Causes and periods of earthquakes in the New Madrid area, Missouri and Arkansas: Science, new ser., vol. 21, 1905, pp. 349-350. . Comparative intensities of the New Madrid, Charleston, and San Francisco earthquakes: Science, new ser., vol. 23, pp. 917-918. . Our greatest earthquakes: Pop. Sei. Monthly, July, 1906, pp. 70-86. Describes phenomena of New Madrid earthquake and compares them to those of Charleston and San Francisco. : — . Earthquakes and the forest: Forestry and Irrigation, vol. 12, 1906, pp. 261-267. Describes the destruction of forests by overthrow, submeigence, etc., by the New Madrid earthquake of 1811-12. Goodspeed Publinhing Co. History of southeastem Missouri, Chicago, 1888, pp. 53-55; 304-307. Quotes accounts by Godfrey Lesieur and EUza Bryan and an unsigned letter in the Louisiana Gazette. Halsteady Hurat. The world on fire, etc.. International Publishing Co., 1902. Quotes previous descriptions of New Madrid earthquake and gives referencee to places of publication. Haywood, John. The natural and aboriginal history of Tennessee up to the first settlement therein by the white people, in the year 1768, Nashville, 1823. Contains an especially graphic account of the New Madrid earthquake in the opening chapter (not included in some other editions of same work). For abstract, see Safford, James M., Geology of Tennessee, pp. 124-125. Hildreth, Scmiael Freacott. American Pioneer, vol. 1, 1842, pp. 139 et seq. (not Seen). . Original contributions to the American Pioneer, Cincinnati, 1844, pp. 34-35. (Reprint from American Pioneer.) Gives a detailed account of the experiences of a party descending the Mitsiflsippi on a flatboat. BIBUOGBAPHY. 113 Höwe, Henry. Historical coUectioiis of the great West, Oincinnati, 1851, vol. 2, pp. 243-246. Gives Flint's description without specific credit (see Flint, Timothy). Humboldt, Alezander von. Cosmos, translated by E. 0. Otto, London, 1849, vol. 1, p.207. Refers to the New Madrid earthquake as one of the few instances of successive shocks throughout long periods in regions remote from volcanoes. James, Edwin. Account of an expedition from Pittsbuigh to the Rocky Mountains underthecommandof Maj. Stephen H. Long, Philadelphia, 1823, 2 vols., vol. 1, p. 272; vol. 2, pp. 325-326. Describes the effect of the earthquake on Indians of the upper Missoiui country and gives an account of one of the later shocks at Cape Girardeau. The nonvol« canic origin of the shocks is affirmed. Latrobe, Charles Joseph. The rambler in North America, second edition, Lon- don, 1836, vol. 1, pp. 107-108. Gives an account of the trip of Oapt. Nicholas Roosevelt, who was taking the New OrUara — ^the first steamer on the river— down the Ohio and Mississippi at the time of the shock. Lesieur, Oodfrey. (The New Madrid earthquake.) Abstract in ^' History of south- eastem Missouri,'' Goodspeed Publishing Co., Chicago, 1888. Abstract in Switzler's ^^ Illustrated history of Missouri from 1541 to 1877," W. F. Switzler, St. Louis, 1879. Abstract in ''The New Madrid earthquake," C. C. Broadhead, Am. Geologist, vol. 30, pp. 79-80. Idnn, Lewis F. (Letter to chairman of committee on Commerce, TT. S. Senate) in ''Gazetteer of the State of Missouri," Alphonso Wetmore, St. Louis, 1837, pp. 131-142. Gives a graphic account of the earthquake and its effects. Lloyd, James T. Lloyd's steamboat directory and disasters on the westem waten, Oincinnati, 1856, p. 325. Gives a general description of the earthquake and some of its associated phe« nomena. . Letter from Cape Girardeau, Louisiana (jazette (Feb. or Mar.?), 1812. Describes phenomena and destruction resulting at Cape Girardeau from the shocks of Jan. 23 and Feb. 7. Lyell, Sir Charles. A second visit to the United States of North America, London, 1849, vol. 2, pp. 228-239. Gives a graphic account of the sinks, fissures, sunk lands, dndned lakes, and prostrated forests in the vicinity of New Madrid. . Principles of geology, 12th ed., London, 1875, vol. 1, pp. 452-453. Abstract of the accoimt appearing in the preceding publication. McGee, W J. A fossil earthquake: Bull. Geol. Soc. America, vol. 4, 1892, pp. 4]ll-413. Gives a detailed account of the uplift and dome west of Reelfoot Lake, the submeiged timber of the region, and the fissures in the face of Chickasaw Bluff. McMurtiie, H. Sketches of Louisville and its environs, Louisville, 1819, pp. 23^255. Gives an appendiz containing complete records of earthquakes by Jared Brooks. See Brooks, Jared. ill, Samuel Latham. A detailed nanative of the earthquakes which occurred on the 16th day of December, 1811: Trans. Lit. and Philos. Soc. New York, vol. 1, pp. 281-307. A compilation of reports of the earthquake from all parts of the United States, especially from the eastem eitles. 114 THE NEW MADRID EARTHQUAKE. MwairlCj John B. Storiee of lÜBsmiri, American Book Co., New York, 1897, pp. 143-150. CoataiDfl a detaüed acooimt of the New Madrid earthquake compiled from eariy descriptions, but without specific credit being given. NnttaU, Thomas. A joumal oi travels into AricanaaB, Philadelphia, 1821, pp. 4^ 47,58. Gives an account ol a viait to the earthquake r^oa in 1818 and notes the ireqnency of shocks at that time. Describes the destruction at New Madrid, Little Piairie (Caruthereville), and Big Prairie. Owen, David Dale. Report of the geological survey in Kentucky, made during the yean 1854 and 1855, Frankfort, 1856, pp. 117-lld. Givea descriptions of earthquake features in the vicinity of Reelfoot Lake. First report of a geological reconnaisBance of the north^n counties of Arkansas, made during the years 1857 and 1858, pp. 31, 203. Notes the occurrence of numerous earth cracks, aand blows, and associated lignite in Poinsett County, and of earth cracks in Greene County, due in part to the shrinking of the underlying aigillaceous strata and in part to the slumping effects of former earthquake action. Pezkins, J. H. Annais of the West, Cincinnati, 1846, pp. 520-524. Quotes in füll the account of Hildreth (see Hildreth, S. P.) and mentions that the first steamboat on the Mississippi (see I^trobe, C. J.) was nearly overwhelmed. Saflord, James M. Geology of Teimessee, Nashville, 1869. Gives an account of earthquake features especially landslides in westem Tennessee and reprints an account of the earthquake and its phenomena. Shaler, Nathaniel Southgate. Earthquakes of the westem United States : Atlantic Monthly, November, 1869, pp. 549-559. Quotes Bradbury's description and gives Drake's list of shocks at Cincinnati. States that centrum moved during continuance of shocks from west of the Mis- sissippi to the mouth of the Wabash River. Mentions fissures 100 feet deep in Obion County, Tenn. Kentucky, Boston, 1885, p. 44. Contains a brief and somewhat inaccurate description of the New Madrid earthquake. Notes on the bald cypress: Memoire Mus. Comp. ZooL, vol. 16, No. 1, 1887, pp. 1-15. Notes death of cypress when water covere knees and quotes Reelfoot Lake, Tenn., as an example, the submergence being due to the New Madrid earth- quake. Many trees are living whoee knees are nearly but not quite submeiged. Shaler, WÜliam« (Letter to S. L. Mitchill): Trans. Lit. and Philos. Soc. New York, vol. 1, pp. 300-302. Gives an account of the experiences of a friend on board a boat on the Mis- sissippi. Shepard, E. M. The New Madrid earthquake: Jour. Geology, vol. 13, 1905, pp. 45-62. Quotes several of the early accounts, describes the earthquake features as now existing, and discusses their relation to artesian conditions, concluding that the sinking was due to undermining by circulating earth waters. Smith, Edward DarrelL On the changes which have taken place in the wells situated in Columbia, S. C, since the earthquakes of 1811-12: Am. Jour. Sei., Ist ser., vol. 1, 1818, pp. 93-95. Failure began alter dry seasons and the year after shock, the loss being only partly restored in the following wet seasons. Author does not aasume earth- quake to be the cause. BIBLIOGBAPHY. 115 Switsdery W. F. Switzler's illustrated history of Missouri from 1541 to 1877 (not Seen), St. Louis, 1879. Reprints a graphic account of the earthquake and earthquake phenomena by Godfrey Lesieur. tJshery F. C. On the elevation of the banks of the Mississippi in 1811: Am. Jour. Sei., Ist ser., vol. 31, 1837, pp. 294-296. Warner, Aug. (Disappearance of Island No. 94), St. Louis Globe-Democrat in March, 1902, quoted by Broadhead, G. C, Am. Geologist, vol. 30, p. 83. Describes the disintegration of the Island during the shock. Wetmore, Alphoxiso. Gazetteer of the State of Missouri, St. Louis, 1837, pp. 131-142. Printe a long letter by Lewis F. Linn, United States Senator, giving a graphic account of the earthquake and its results and gives a few original remarks stating that earthquakes were still feit every two weeks or so. INDEX. A. Page. Aclmowledgments 7-9 Acti vity , periods of , by houn 36-36 periodsof, by wedcs 33-34 Area affected, eztent of. 7,10-17 Arkansas, eflectsofshocks in 21 Atmosphere, phenomena of. 44-47 Aadubon, John J., on natura of tfae vibra- tions 31 B. Baily, C. B., acknowledgments to 8 Bayoa Channels and lakes, fonns of, flgure showing 67 Barometric pressure, distribution of shoclcs withreferenoeto 36-^ Baiton, William, on an earthquake in 1796. . . 11 Beechwell, Mo., fissure at, flgure showing. . . 54 Bibliograidiy 111-115 Big Lake, description of 71-72 Blytheville dorne, ageof. 12 Boston, tremors feit in 17 Bradbury, John, on agitation of water sur- laces ••••••...................... ov on caving of river banks 92 on nature of the vibrations 32 Bringier, L., on cause of the earthquake 103 on extrusion 76, 86 on nature of the vibrations 31 on sand blows 82 on sinks 88 on Splitting of trees 95 Broadhead, O. C, on caving of Island No. 94. . 43 Brooks, Jared, on agitation of water suitaces . . 92 onintensitiesoftheshocks 33 onnatureof the vibrations 32-33 reoord kept by 22-26 Bryan, Eliza, on agitation of water surfEUXS. . 90 Buildings, eflect on 99-101 C. Canada, tremors feit in 17 Cape Girardeau, Mo., destruction at 100 Garrigan, John, on the earthquake in South Carolina and Oeorgia 30-31 Caruthersville. 8u Little Prairie. Carthage, Tenn., eflects of the shocks at . . . . 21, 100 Casseday , Ben, on nature of the vibrations. . . 31 Cause of earthquake, evidenoe as to 103-104, 105 volcanic theory of 102-104 Caving of river banks, reoords of. 92-93 Centrum of earthquake, location of . . 14-16, 104-105 Charieston, S. C, eflects of the shocks at 30 Chickasaw Biufls, landslides on 59-60 Christian County, Ey., eflects of the shocks in 21 Cincinnati, Ohio, destruction at 101 recordof the shocks at 26-27 Page. Cirdeville, Ohio, effects of the shocks at 28 Qarksville, Tenn., eflects of the shocks at. . 21, 100 Goal. i9eeLignite. Cole, P. H., on changes in a spring 95 Columbia, 8. C, eflects of the shocks at 30 Columbia, Tenn., darkness at 45 eflects of the shocks at 21 D. Darkness, {»evalenceof 44-45 Dillard, A. N., on Splitting of trees 95-96 on sunk lajids 70 Dome, Biytheville, description of 63-64 Blytheville, records of 62 Little River, description of 64 reoords of. 62-63 Tiptonville, description of 63 records of. 62 Doming, cause of 64,106-109 possible cause of , diagiam showing 106 reoords of. 62-64 signiflcance of 105-106 Drake, Daniel, on nature of the vibrations.. 32 on previous reoorded earthquakes 11-12 record kept by 26-27 E. Earthquake, activity of, by hours of the day, diagram showing 36 activity of, by intensities, diagram show- ing 35 by phases of the moon, diagram show- ing 38 by weeks, diagram showing 34 Earthquake disturbances, extent of, map showing 8 Earthquake features of the New Madrid dis- trict, map showing In pocket Southwest of Lake St. Francis, map show- ing 50 Earthquakes, contemporaneous, record of. .. 109 earli^, in the Mississippi valley, geologic evidence of. 12-13 in the Mississippi valley, reoords of . . 11-13 future, possibilities of 109-110 Extrusion, cause of 85-87 material brought out by 76-77, 78 records of 76,78 F. Fault trenches, cause of 87 diagram showing 48 plate showing 12 Faults, location of 58 cause of 50 117 •.■■V.'/.'.^A V.V IW •mm •^L'iam\im\n at 17 Little River dorne, age of 12 Lloyd, J. T., on agitation of water surfaces . . 80 Louisville, Ky., destructk>n at iDO distribution of shocks at 33-40 record of the shocks at 22-26 Lyell, Charles, on dead trees 08 on earlier earthquakes 12 onsinks 87 on sunk lands 70 Visit of,to8oeneof earthquake 10 IC. McQee, W J, description of Reelfoot Lake by . 73 ontheTiptonviUedome 63 McMurtrie, Henry, records published by 22 Mississippi River, changes in tha fthonnAi of , Sketch showing 93 Mltchlll, S. L., on effects of the shocks in the Atlantic coastal piain 20-31 Moon, phases of, distribution of shocks with referenoeto 37-39 Musk oz, cranium of, extrusion of 77 N. Natebec, Miss., effects of the ähocks at 21, 100 Navigation, effect of earthquake on 02-04 New, Anthony, on shocks at Red Bank 28 New Madrid, layout of 00-100 populatwn of 100 New Orleans, tremors feit in 17 New Orleans (steamer) voyage of. 02-93 Noises, reoords of 101 Norfolk, Va., effects of the shocks at 29 Number of the shocks, table showing 17-20 Nuttall, Thomas . on cause of the earthquake. . 103 O. Observers, educated, kKatwns of O-IO Oders, production of 45-46 Omori, observations of 39 INDEX. 119 P. Page. Periodsofactivity 3»-34 PineTilte, 8. C. , eflects of the shocks at 30 Piney River, Tenn., eflects of the shocks at. . 21 Pittsburgh, Pa., eflects of the shocks at 28 Portsmouth, Va., eflects of the shocks at — 20 R. Raleigh, N. C, eflects of the shocks at 29 Red Bank, Ind., eflects of the shocks at . 27-28, 100 Red River Settlements, tremors feit in 17 Reelfoot Lake, description of 73-74 landslides near 61 submerged ch anneis in, map sho wing — 69 section across, figure showing 74 Kichmond, Va., eflects of the shocks at 29 Riddick, Mr., on light glows 46 Roosevelt, Capt., voyage of 92-93 S. St. Genevieve, Mo., eflects of the shocks at. . 22 St. Ix>uis, Mo., eflects of the shocks at 22, 100 Sand, extrusion of 76.77,80-85 Sand blows, coolesceat, plate showing 80 distribution of 80-83 fissures of 57 original appearance of, plate showing 76 present appearance of, plate showing .... 80 proflies of, Agare showing 79 nature of 79-80 Situation of 80 Sand scatters, nature of 85 Sand sloughs, cause of 85, 87 distribution of. 84-85 fissures of 57 nature of 83 Situation of 83-84 Savannah, Oa., efliBCts of the shocks at 30 Sawyers, disturbanoe of 94 Sevcrity of the shocks, testimony to 17 Shaler, N. S., on character of the earthquakc. 7 on Position of the centrum 14, 15 Shaler, William, onagitationof water surfaccs 90-91 Shepard, £. M., acknowledgments to 7-9 on later extrusion of water 78 publication by 9 Sinks, description of. 87-88 production of. 88-89 Smoke-house inddent, aocount of 55 Snags, disturbance of 94 South Carolina, destruction in 101 Springs, changes In 95 movements of ground water producing, diagram showing 78 Summary of earthquake 10-11 Sunk lands, cause of. 74-75 distribution of. 68-74 Page. Sunk lands, form and trend of. 66-67 form of Channels in, flgure showing 66 nature of 64-65 ilver swamp, type of, plate showing 64 stream crosslng, section of, flgure showing. 65 subsidenoe of, amount of 67-68 tyi)esof 65 water weeds in, plate showing 64 young growth in, plate showing 70 T. Thrust, lateral, eflect of 107-109 Time of shocks, reports of 13-14 Tiptonville dome, age of. 12 Trees, dead, occurrence of 98 elevated, plate showing 72 in sunk lands 70-71, 72, 73-74 kiUed by deposits of sand, plate showing. 84 kllled by submergenoe, plate showing 60, 72 figure showing 71 Splitting of 95 tilted by earthquake, plate showing 60 tilted on landslides 60-61 uplifted, occurrence of 98 with double sets of roots, plate showing. . 84 See also Forests. Turner, George, on an earthquake in 1795 11 r. Undermining. eflects of 87-89, 107 Uplifts, cause of 106-109 records of 62-63 V. Vaiwrs, escape of 45-46 Vamey River, Channel of , plate showing 70 Vibrations, destruction of forests by 97 directionof 40-42 nature of 31-33 Vincennes, Ind . , eflects of the shocks at 28 Volcanic action, oontemporaneous, record of . . 109 W. Washington, D. C, eflects of the shocks at. . . .17, 29 Water, cxtruded, records of 76,78 extruded, temperature of 77-78 surface of , agitatlon of 89-02 Weather, oondition of , before earthquake 13 relation of the earthquake to 39-40 Whistler, Capt. William, on an earthquake in 1804 11 Witherall, James, on eflects of the shocks at Detroit 28 Wood ward, A. B., on eflects of the shocks in Michigan and Canada 28 Wreckage, floating 94 O * ü. S. GOVERNMENT PRINTING OFFICE : 1992 - 316-859 : QL 1 r Twiir f. i . . tu- y. 1 ^i- .>• H» •^■. ♦ «^ ■ • -« ■ ' » « k *« '• • i^ \\ i DATEDUE » tj n ^^^^ STANFORD UNIVERSITY UBRARIES STANFORD, CAUFORNIA 94305-6004 STANFORD UNIVERSrTY UBRARIES STANFORD, CAUFORNIA 94305-6004 nm