1920.] Hogben.—Interpretation o.f a Typical Seismogram. 109 
(6) The x waves may be explained in several ways—as S waves four 
times reflected (SR 4 ), for instance. Probably, like the L waves, they are a 
compound phase. 
(8), (9), (10): We have still to account for L 2 , L 3 , L 4 . There are other 
phases in the “tail” of the record, some of which are due to waves travel¬ 
ling along the major arc ; but L 2 , L 3 , L 4 are evidently minor arc waves. 
It appears to me that the L 2 waves are most easily explained by supposing 
them to be repeatedly reflected between the earth’s surface and the surface 
of an inner shell—say, II; similarly, we can explain L 3 and L 4 by repeated 
reflexions between the earth’s surface and the surfaces of other shells—say, 
III and IY. The velocities of propagation of L 2 , L 3 , and L 4 are about 
3*6, 3-4, and 3*0 kilometres per second respectively; it is evident, therefore, 
that the L 2 waves must go deeper than the L 3 waves, and these deeper 
than the L 4 —that is, the reflecting surface II is deeper than III, and III 
than IY. The total length of the paths of several series of the L 2 waves 
will differ so little that they will reinforce one another, and give this part 
of the seismogram its marked character. So also the L 3 and L 4 waves. 
(The existence of such shells is, of course, a hypothesis to be tested.) 
If P and S are accurately determined for a large number of earthquakes 
for different epicentric distances, we can determine the paths of the waves 
of various phases, and hence make reasonable inferences as to the elasticity, 
rigidity, and density of the rocks at different depths, even, possibly, as far 
as the earth’s centre. But for this a high degree of accuracy is necessary— 
accuracy which is to some extent attained in the Wiechert instruments, 
and still more in the excellent seismographs invented by the Russian Prince 
Galitzin, but reached not at all by the Milne machines. 
The seismograph should be sensitive and have a high magnifying-power, 
otherwise it may fail to record many of the phases of an earthquake, 
especially the P waves, for which, in that case, PB 4 may be mistaken. 
Again, the Milne horizontal pendulums, when started by an earthquake, 
quickly acquire their own natural period ; a good instrument would be 
properly damped, giving only the motion impressed on it (really on the 
earth’s surface) by the earthquake. The clock-time, too, should be accu¬ 
rately kept and marked on the seismograms, although this is only a matter 
of ordinary care in an observatory. 
I am glad to say that the New Zealand Government has agreed to install 
at the Hector. Observatory a Milne-Shaw seismograph — the instrument 
recommended by the B.A.A.S. Committee where it is not possible to set up 
Galitzin machines. This will put our seismological work here upon a good 
footing, which will be comparable with that of other first-class British 
observatories when another Milne-Shaw is set up at right angles to the first, 
as well as a vertical seismograph of good type. 
Apia, Samoa, is well equipped with Wiechert instruments. It is to be 
hoped that the New Zealand Government will see its way to maintain this 
observatory, which is so important a link between the seismological observa¬ 
tories of New Zealand, Australia, and eastern Asia, and those of North and 
South America. In fact, it would be difficult to name an observatory whose 
geographical situation gives it more importance than Samoa. It is highly 
important, too, that the work should be of first-class quality. This implies 
well-trained assistants who are enthusiasts in their work—which tempts 
me to comment upon a cablegram received here when the British Associa¬ 
tion was in session. That message stated that all British stations should 
