ON SEFSMOLOGICAL INVESTIGATION. 83 



times that of tlie instruments employed at the stations in Britain, from 

 which it might be inferred that very minute ^preliminary tremors might 

 be recorded, and therefore earlier commencements of motion Ije calculated 

 for these Continental stations than would obtain in Britain. 



With the assumption that the greatest difference in time that could 

 exist between the commencement of motion at these two groups of 

 stations is four minutes, the comparison of fifty-six records common to 

 Germany and Britain leads to the following : — ■ 



In twenty-four instances the difference in the times of commence- 

 ments does not exceed the four-minute limit. These in the Shide register 

 correspond to numbers 581, 58+, 585, 586, 588, 590, 595, 606, 614, 616, 

 619, 6196, 625, 627, 636, 641, 642c', 644, 653, 658, 661, 662, 663, 665. 



The remaining thirty-two instances where this limit has been exceeded 

 refer to twenty-one mere thickenings of tlie trace and eleven to earthquakes 

 with moderate amplitudes. These thirty-two instances may be divided into 

 two groups, there being twenty-three cases where the British records are 

 late relatively to those noted in Germany, and nine when the German 

 records fall behind those obtained in Britain. The British records, which 

 are late, are numbers 578, 580, 583, 597, 598, 600, 6006, 6066, 611, 6136, 

 617, 618, 6226, 624, 633, 639, 640, which are all minute thickenings on the 

 trace, and 589, 592, 599, 609, 612, and 659, which are well-defined records. 



The German records, which are late, are numbers 576, 582, 610, 

 which, as noted in Britain, are small, and numbers 572, 593, 601c, 607, 

 626, 642, which are large or fairly large disturbances. The number of 

 disturbances as recorded in Germany with too late commencements, oddly 

 enough, is exactly the same as recorded in Britain. 



The conclusions to which these comparisons point are : — 



1 . For recording small tremors which do not extend over great areas 

 the Rebeur-Ehlert pendulum, as installed at Strassburg, possesses advan- 

 tages over the Milne horizontal pendulum as installed at stations 

 co-operating with the British Association. 



2. For recording the commencements and, it may be added, other 

 phases of earthquake motion which affect the world as a whole the 

 accuracy of the records from both types of instruments is practically 

 identical. 



In connection with these conclusions it must be pointed out the 

 fineness of the trace obtainable with the British Association type of 

 instrument partly compensates for its comparative want of sensibility. 

 The particular sensibility given to it is one that is obtainable at a variety 

 of stations. Were this increased, which is easily done by raising its 

 period from sixteen to twenty, or even forty, seconds, when it would be 

 more responsive to tremors, then at many stations it would be found 

 that diurnal and other wanderings, together with air tremors, would 

 seriously interfere with the recording of earthquakes. Instruments of 

 the Rebeur-Ehlert type, with large multiplication, not only consume what 

 for many would be a prohibitive quantity of photographic paper, but, as 

 for example at Trieste and Kremsmiinster, they are frequently recording 

 movements which are not required. 



q2 



