496 



8CIENCE. 



[Vol. IX., No. 224 



cannot, of course, deternaine, though it may be 

 possible to obtain some notion of its nnost general 

 features if the data are sufficient. 



This method of computing- the depth of a seis- 

 mic focus is here proposed for the first time. 

 The method employed hy Mallet, which consists 

 in finding the angle of emergence of a wave- 

 front from the earth l)y studying the configura- 

 tion of cracks in buildings, is believed to be 

 pretty nearly valueless by all seismologists. There 

 is no definite angle of emergence, of the nature 

 he contemplates, disclosed at the surface. Cer- 

 tainly in Charleston there was nothing of the kind 

 to be found. The method employed by Seebach 

 is sound in theory, but it requires such extreme 

 accuracy of time-determinations that very small 



no means of determining ; but we do not be- 

 lieve that it would be so affected to any great 

 extent in such a region as South Carolina. 

 Being independent of any absolute measures 

 either of the surface intensity or of the total 

 energy of the shock, the greatest difficulty of all 

 is at once eliminated. Our own opinion of this 

 method is, that it is incapable alike of very great 

 precision and of very great errors. 



Probably the first thought occurring to any 

 one examining this method will be that the de- 

 termination of the two required points would be 

 liable to very large errors. But, if he will exam- 

 ine the varying values of the ordinates of the 

 curve corresponding to varying values of the ab- 

 scissas and of the depth, we think he will be sat- 



S0I9 IB ir ti/SK-tMZII 10 9 & 7 G 5^9 Z / < 



N.W. 



Fig. 5.— Intensity curve op Charleston earthquake. 



errors of time give very large errors in the result. 

 Our own method consists in finding tw^o points 

 on opposite sides of the seismic vertical, at which 

 the changes in seismic action along a given line 

 are most strongly marked. These points ought 

 to be indicated in powerful earthquakes with a fair 

 approach to precision, and the probable errors of 

 determination should not usually exceed one or 

 two tenths of the distance between the two points. 

 The feebler the shock, however, the less is the 

 degree of precision to be expected. Whatever 

 may be the errors in the estimate of this dis- 

 tance, the resulting error in the computed depth 

 is smaller than the error of observation in the 

 ratio of the square root of three to two. How 

 much the estiuiate may be vitiated by want of 

 homogeneity in the superficial strata, vs^e have 



isfied that the limits within which each of the 

 two points of inflexion must fall cannot be wide 

 apart, and that an error in the determination of 

 the base-line greater than two-tenths of its esti- 

 mated length would in such a country as Caro- 

 lina be very improbable. It will appear that the 

 relations of these variables are such as to restrict 

 the locus within which the desired points are to 

 be found to a very narrow annulus around the 

 epicentrum. We believe the method will improve 

 upon acquaintance. 



We have endeavored to apply our method of 

 computing the depth of the focus to other earth- 

 quakes, but have found difficulty in obtaining 

 any thing more than very general results, such as 

 the following. The depth of the Charleston earth- 

 quake was relatively great ; and we find reason 



