THE RECENT ADVANCE IN SEISMOLOGY 399 
have an arcual path similar to, but deeper than, the “‘large waves,” 
which follow them in time and are, as all agree, surface waves.* 
The velocity of these first waves, reckoned on this basis, would be 
about 14 kilometers per second instead of to. Perhaps the strongest 
argument against this view is found in the known constants of surface 
rocks, which do not permit such a velocity. (See below, p. 400.) 
It seems likely that the value assigned for the velocity of the direct 
waves may be subsequently considerably modified, since it has been 
largely based upon the records of the light Milne pendulums. Some 
sacrifice the pioneer must always make, and the standard Milne 
instrument suffers in comparison with the later German, Japanese, 
and Italian types, not only because of its lightness, but because the 
expense of photographic paper has necessitated a slow movement 
of the feeding-drum and a resulting contracted scale of the diagram. 
At the Batavia station a Milne and a Rebeur instrument have been 
installed side by side, and a comparison of the records now shows 
for the first time that the registration of shocks begins from one to 
ten minutes earlier upon the Rebeur pendulum.? The “ preliminary 
tremors” of the Milne instrument may belong in the second phase 
of the records from more sensitive instruments. 
For origins less than 1,000 kilometers distant such tremors do not 
appear in the seismogram, and it is supposed that they are combined 
with the large waves and reach the station at a speed of about 3.3 
kilometers per second, with little doubt because of the lower rock 
densities which are traversed along the shorter and hence “‘crustal”’ 
chords. 
The Earthquake Investigation Committee of Japan in 1894 insti- 
tuted, at the suggestion of Professors Sekiya and Omori, a system of 
triangulation involving the use of four stations provided with exactly 
similar instruments and connected by telegraph to convey uniform ticks 
from a chromometer. The distances between the stations varied 
from 2.29 to 10.86 kilometers. All instruments being started by 
the same earthquake, the recognition of special marked vibrations 
t Baron Dairoku Kikuchi, “Recent Seismological Investigations in Japan,” Pub. 
E.I. C. (Foreign Languages’, No. 19, 1904, p. 61. 
2E. Rudolph, “Ostasiatischer Erdbebenkatalog”’ (1904), Gerland’s Beitrége 
sur Geophysik, Vol. VIII, 1906 (1907), pp. 113-217. 
