498 



SCIENCE. 



[Vol. IX., ISO. 224 



record refers ; and this difiSculty is a very serious 

 one. At Summerville the first shock came almost 

 like an explosion. Before people had time to 

 think, they were pitched about like ten-pins. At 

 Charleston there was a perceptible interval, esti- 

 mated at from five to eight seconds, from the first 

 note of warning to the maximum of the great 

 shock. At Savannah (90 miles distant) the inter- 

 val from the beginning to the first maximum was 

 considerably longer, probably ten to twelve sec- 

 onds ; at Augusta (115 miles) the interval was still 

 greater : and, generally speaking, the greater the 

 distance, the more the phenomena were ' long 

 drawn out.' The duration of the eai'thquake at 

 Charleston will probably never be known with ac- 

 curacy, but the general testimony ranges between 

 fifty and ninety seconds. At Washington (450 miles) 

 Professor Newcomb, with his watch in his hand, ob- 

 served a duration of perceptible tremors with two 

 maxima lasting about five and one-half minutes. 

 Professor Carpmael's magnetographs recorded the 

 disturbance, and he interprets their photographic 

 traces as showing a duration of about four minutes. 

 Mr. G. W. Holstein of Belvidere, N.J., gives five 

 minutes very nearly as the observed duration. 

 From other localities come well-attested observa- 

 tions showing durations of several minutes, though 

 few of these pretend to give the whole time with 

 any accuracy. This progressive lengthening of the 

 shocks is a well-marked feature of the testimony. 

 The explanation suggests itself at once. The elas- 

 tic modulus of compression being greater than 

 that of distortion, the speed of the normal waves 

 is the greater, while the waves of distortion lag 

 behind. 



It is obvious that the phase which it is desired 

 to observe should be the arrival of the first im- 

 pulses, but the great duration of the tremors has 

 left much doubt on this point. Stopped clocks 

 were plentiful all over the country, but at what 

 phase of the earthquake did they stop? So great, 

 indeed, are the uncertainties on this point, that 

 the observations of intelligent men, with watches 

 in their hands, measuring a part of the shock and 

 estimating the beginning, are in most cases to be 

 preferred to stopped clocks, even if we knew 

 with certainty that the clocks had been accurate 

 to the second. It matters little how we twist and 

 turn the time data : the smallest estimate we can 

 put upon the speed of propagation must prove to 

 be a great surprise to seismologists. 



The time at Charleston of the occurrence of 

 the main shock has been fixed at 9.51.10 P.M., 

 75th meridian, or eastern standard time (all times 

 in this paper, unless otherwise specified, are re- 

 duced to that meridian). The uncertainty does 

 not exceed ten seconds. The beginning of the 



first tremors at Charleston was from six to eight 

 seconds earlier. The time at Summerville was 

 probably less than four seconds earlier than 

 Charleston. For all localities within two hundred 

 miles, the time observations are of little value. 

 So swiftly did the waves travel, that a small 

 error in the time record gives a very large uncer- 

 tainty in the resulting speed. 



The nearest point which yields a valuable rec- 

 ord is Wytheville, Va. (286 miles).' Mr. Howard 

 Shriver was sitting at a transit instrument, wait- 

 ing for the passage of a star, and at once noted the 

 time at 9.53.37 (reduced to 75th meridian), giving 

 a speed of about 3.3 miles (5,300 metres) per sec- 

 ond. There is some slight uncertainty about the 

 precise phase of the shock corresponding to the 

 observation. 



The signal service observer at Chattanooga (332 

 miles) gives only the nearest minute for the prin- 

 cipal shock at 9.53, corresponding to a speed of 

 3.02 miles per second, or 4,860 metres. 



The best observation in our possession is that of 

 Prof. Simon Newcomb himself, at Washington 

 (450 miles), who gives the time of the beginning 

 of the shock at 9.53.20, with an uncertainty not 

 greatly exceeding ten seconds. The resulting 

 speed is 3.46 miles per second, or 5,570 metres. 



From Baltimore (486 miles) Mr. Richard Ran- 

 dolph, C.E., reports a very intelligent and care- 

 fully verified observation of 9.53.20 as the begin- 

 ning of the shock, — exactly Professor Newcomb's 

 time for Washington, giving a speed of 3.74 miles, 

 or 6,000 metres, per second. 



At Atlantic City, N.J. (552 miles), a large pen- 

 dulum-clock in the Fothergill House stopped at 

 9.54 very nearly. If this may be taken to be the 

 beginning of the shock, the speed would be 3.26 

 miles per second, or 5,250 metres. 



George Wolf Holstein, Belvidere, N.J. (632 

 miles), gives 9.54 for the beginning of the shock, 

 and 9.59 for the end, and compared his watch 

 next morning with the time of the Pennsylvania 

 railroad. The gradual and uncertain character of 

 the beginning and end would not adniit of precise 

 determination to seconds. The speed, taking 9.54 

 for the beginning, would be 3.66 miles, or 5,900 

 metres. 



From New York City (645 miles) and its sub- 

 urban towns and cities come many reports, all of 

 which give either 9.54 or 9.55 as the nearest min- 

 utes. If we take as a mean 9.54.25 at New York 

 and Brooklyn for the beginning of the shock, the 

 speed would be 3.31 miles, or 5,330 metres. 



1 The distances are measured somewhat hastily with a 

 scale upon the war department map of the United States, 

 taking the greater epicentrum 16^ miles north-west of 

 Charleston as the starting-point. 



