Januabt 22, 1909] 



SCIENCE 



133 



to the future, and especially to such 

 adaptations of seismographie appliances 

 and methods as we may confidently antici- 

 pate. Feeble tremors, ascribed to minute 

 crepitations of the crust, have already 

 been made audible by means of the micro- 

 phone, so that the ear could be applied to a 

 sort of seismic stethoscope; and the next 

 step will perhaps be to construct a seis- 

 mograph of such delicacy as to record 

 these minute vibrations, and install it 

 where it will be undisturbed by tremors of 

 artificial origin. 



Summarizing briefly, many of the mal- 

 loseismic districts or areas of earthquake 

 danger are known from records of past 

 experience, and others are being recog- 

 nized by physiographic characters. Within 

 them are tracts of special instability be- 

 cause of the incoherence of the underlying 

 formation, and these can be both char- 

 acterized in general terms and locally 

 mapped. The general relations of danger 

 to place are so well understood that an 

 early solution of their outstanding prob- 

 lems may be assumed. Of the relations 

 of danger to time much less is known and 

 there is less promise for the immediate 

 future. The hypothesis of rhythmic re- 

 currence has no sure support from obser- 

 vation, and is not in working order for 

 either large or small areas. Its corollary 

 of local immunity after local disaster is 

 more alluring than safe. The allied hypo- 

 thesis of alternation between parts of a 

 district is being tested by a great example, 

 but the verdict belongs to the future. 

 The hypothesis of precipitation by acces- 

 sory forces which are in large part 

 periodic and foreknown, has a good status 

 and is being developed on the statistical 

 side. It promises to make the date of pre- 

 diction more precise if ever the approxi- 

 mate time shall be achieved by other 



means. The hypothesis of an intelligible 

 prelude has barely been broached and the 

 means to test it are not yet in sight. In a 

 word the determination of danger districts 

 and danger spots belongs to the past, the 

 present and the near future; the determi- 

 nation of times of danger belongs to the 

 indefinite future. The one lies largely 

 within the domain of accomplishment; the 

 other stiU lingers in that of endeavor and 

 hope. 



We may congratulate ourselves that it 

 is not the place factor which lags behind, 

 for knowledge of place has far more prac- 

 tical value than knowledge of time. In 

 fact I see little practical value in any 

 quality of time precision attainable along 

 lines of achievement now seen to be open. 

 Suppose, for example, that a prediction 

 based on rhythm or alternation should 

 indicate an earthquake as due in a certain 

 year, and that tides should be recognized 

 as the most potent accessory cause; then 

 for several days each month, and possibly 

 for many months, expectation would be 

 tense, and the cost in anticipatory terror 

 would be great. Or suppose that prelude 

 phenomena should be found to afford real 

 warning; the forecaster on duty would 

 still have to deal in probabilities, and 

 when in doubt would often sound vain 

 warnings, in the conscientious effox't to es- 

 cape the greater error of omission at the 

 critical time— and again nervous strain 

 would be wasted. And even if warning 

 were definite, timely and infallible, so 

 that peril of life could be altogether 

 avoided, property peril would still remain 

 unless construction had been earthquake- 

 proof. If, on the other hand, the places 

 of peril are definitely known, even though 

 the dates are indefinite, wise construction 

 will take all necessary precautions, and the 

 earthquake-proof house not only will in- 

 sure itself but will practically insure its 

 inmates. 



