June 17, 1887.] 



■SCIENCE. 



585 



plished by the patient and laborious investigation 

 of a mass of evidence, much of which is unsatisfac- 

 tory, and not a little of it untrustworthy. That the 

 paper contains so mtich that is valuable and interest- 

 ing is greatly to the credit of its authors ; and its 

 real importance as the most, indeed the only, elab- 

 orate discussion of that interesting seismic event 

 which has thus far apjDeared, renders a careful ex- 

 amination of its methods and conclusions extremely 

 desirable. I wish to remark upon a few points, con- 

 cerning which I am compelled to dissent from the 

 views expressed in the paper. 



Great labor has evidently been expended on the 

 construction of the isoseismal chart; and doubtless 

 all has been done that can be, with the uncertain 

 data available. Two serious but well-recognized 

 difficulties are met with in attempting the construc- 

 tion of ' intensity ' curves : one is the variability 

 and inconsistency of the physical evidences of dis- 

 turbance, and the other is the unreliability of human 

 testimony as to its extent. In earthquakes of much 

 violence a considerable area near the origin may pre- 

 sent evidence which outlasts the disturbance itself, 

 such as overthrown or damaged buildings, chimneys 

 overturned, monuments displaced, etc., and which, 

 therefore, may be sttidied at leisure. A little experi- 

 ence in the examination of this sort of evidence 

 proves conclusively that a given result is an ex- 

 tremely complex function of a large number of ' in- 

 dependent variables,' most of which, unfortunately, 

 are and must be unknown. It thus becomes difficult 

 to determine the ratio between the varying magni- 

 tudes of any one of these variables, so largely is the 

 visible result influenced by the others. 



The available source of information consists gen- 

 erally of the effects of the disturbance upon struc- 

 tures of various kinds. Nothing can be more con- 

 flicting than the results of siich observations, even on 

 areas so small that it seems impossible to admit that 

 differences in actual earth-movement have existed. 

 Within a hundred feet of each other will be found 

 buildings nearly destroyed, and buildings, apparently 

 similar in construction, almost uninjured. Here a 

 monument or shaft is overthrown ; and there, a few 

 feet away, another on a mxich less stable foundation 

 is undistxirbed. In a room in which heavy bookcases 

 have been dashed upon the floor, and the furniture 

 generally wrecked, delicate ornaments still rest upon 

 the mantelpiece, and, without crack or scratch, 

 seem to deny all possibility of violent motion. In 

 short, one is forced to the conclusion that the char- 

 acter and amount of destruction caused by an earth- 

 quake depend largely on circumstances other than 

 the motion of the earth-particle. An earthquake 

 must be studied in the light of what it has failed to 

 do, as well as of what it has done ; and much con- 

 sideration should be given to what might have hap- 

 pened but did not. 



If such widely different effects can be produced 

 by earth-movements which must be practically the 

 same, it is clear that they cannot be very accurate 

 measures of the intensity of seismic disturbances. 

 In a general way, and if extended over an area which 

 includes decided changes in the extent of surface 

 destruction, such observations are extremely useful 

 as indicating zones of unequal disturbance ; and 

 especially so, as, in the absence of instrumental 

 records, they furnish about all the available facts. 



The nature of the data furnished by the careful 

 and conscientious survey of Mr. Sloan is not stated ; 



but it is perfectly safe to say, that, whatever it may 

 be, Messrs. Dutton and Havden have made the most 

 of it. 



Without intending any special criticism upon the 

 method of treatment adopted, I desire to call atten- 

 tion to the uncertainty, which seems to be great, in 

 the construction of equal intensity-curves with any 

 attempt at precision in form or position, when they 

 are based upon observations of such physical dis- 

 turbances as are referred to above. 



If such records of the disturbance as are left by the 

 earthquake itself are of doubtful and uncertain value, 

 still more so must be the data resting entirely upon 

 the testimony of observers of transient phenomena. 

 It is by no means uncommon for two persons sitting 

 in the same room, and disturbed by the same moder- 

 ate earthquake, to differ decidedly in their estimate 

 of its intensity. 



In two differently constructed or differently situ- 

 ated buildings near to each other the difference is 

 very great. Nor will it do to dej^end upon the dis- 

 turbance of movable objects, such as swinging- 

 lamps, etc. Very much depends upon the character 

 of the movement, — as to whether the motion is prin- 

 cipally horizontal or vertical, the period long or 

 short, and the synchronism of that period with that 

 of the moving object. Innumerable illustrations of 

 this fact might be given. Disturbances of unusual- 

 ly large amplitude but long period are sometimes 

 scarcely perceptible to the observer. Professor 

 Milne recorded a disturbance in Tokj^o on Nov. 

 23, 1884, of which he says, " Whilst standing up, it 

 was with difficulty perceptible. In the same room, 

 however, those who were seated felt it distinctly. It 

 made a lamp six feet long swing through an arc 

 about six inches." 



In 1881 an earthqviake occurred at Sapporo 

 (Japan), concerning which the observer made this 

 note: "Wire of hanging-lamp four feet long de- 

 scribed an arc of twelve inches ; not personally ob- 

 served ; was walking on the street, and nothing was 

 noticed." 



Besides the physical environment of the observer, 

 his physiological and psychological peculiarities 

 largely control his estimate of the extent of the dis- 

 turbance. 



In the collection of information by means of dis- 

 tributed circulars, it is impossible to avoid these 

 difficulties, and to obtain any thing like a fair esti- 

 mate of the character of the phenomenon, especially 

 as most observers are inexperienced. A circular 

 sent to a village is generally likely to find its way 

 into the hands of the particular inhabitant who can 

 give the most startling account of what he saw and 

 felt, and who was naturally most thoroughly frigh- 

 tened. 



It appears, therefore, that as far as the value of 

 the collected data is concerned, the great area dis- 

 turbed by this earthquake might be divided into 

 three zones. The first is small, surrounding and in- 

 cltiding the epicentrum, and the visible evidences 

 of the intensity of the shock were carefully studied 

 by a sagacious observer within a few weeks of its 

 occurrence. The second consists of the remainder 

 of the area within two or three hundred miles of the 

 epicentral tract, throughout which, though to a con- 

 stantly diminishing extent, overthrown chimneys, 

 displaced shafts, cracked walls, etc., remained as ex- 

 ponents of the character: and magnitude of the dis- 

 turbance. From this region, however, evidence 



