ON SEISMOLOGICAL INVESTIGATION. 71 



light reflected from a mirror liung by a bifilar attachment between the 

 two pointers. Subsequently the record was made mechanically. 



With the first installation 1 mm. deflection = 0"u, and in the second 

 6"-0. 



Although several large earthquakes took place, no record was 

 obtained. 



About the same time Dr. Wilhelm Schliitter (see his 'Inaugural 

 Dissertation,' Gottingen, 1901) experimented with a balance form of 

 clinometer. The records were photographic, but his photograms failed to 

 give any trace of twenty earthquakes recorded by seismographs. 



VIII. Exjjeriments with a Vertical S^yring Seismograph, 



With the object of at least detecting the vertical component of the 

 large waves which accompany unfelt earthquakes, in March 1901 I 

 suspended from the wall of my laboratory an ordinary spiral spring, 



1 inch in diameter, which, under the influence of a load of 1 lb. 8 ozi and 

 its own weight, was 3 feet 5 inches in length. Its period was then 



2 seconds. By the rising or falling of the weight a small mirror was 

 caused to rotate, which displaced a spot of light it reflected upon a 

 moving photographic sui'face. 



'The earthquake of October 9, 1900, caused ripples on the photogram 

 each about O'o mm. in range, which would correspond to a change that 

 might have been produced by increasing and decreasing the load by 

 -jlxo P^-rt of itself. The period of motion was approximately 6-5 minutes, 

 which corresponded with the period of maxima in the large waves as in 

 an ordinary seismogram. 



The Venezuela earthquake of October 29 gave deflections of half the 

 above, and with periods of about 7 minutes. Other earthquakes caused 

 somewhat similar movements, but usually nothing more than slight blurs 

 upon the photographic traces were to be seen. 



The records from the clinometer indicate that earth tilting has not 

 been measurable by the instrument employed, whilst the records from the 

 spiral spring show that there is a possibility that vertical motion may 

 exist, but if it does it is exceedingly minute.' 



IX. On the Nature of Earthquake Movement as recorded at a great 

 distance from its origin. 



In an article in ' Nature,' January 2, 1902, p. 202, after epitomisihg 

 the various observations which have been made in connection with the 

 large waves of earthquakes, it is said that 



' The general inference is that the large waves due to earthquakes 

 originating at a distance, whether they are surface waves or mass waves^ 

 actuate horizontal pendulums by horizontal displacements of the ground, 

 rather than by the tilting of the same,' 



Observations which support this view are as follows : — 



1. Clinometers have hitherto failed to detect any tilting efiects. 



2. If it is assumed that the records of horizontal pendulums give 

 angular values for tilting, and from the period of the waves causing these 

 tiltings and the velocity with which these waves are propagated on the 

 assumption of simple harmonic motion we calculate their length, we have 



