286 REPORTS ON THE STATE OF SCIENCE, ETC. 
Periodicity. 
A periodicity near 21m. was mentioned in the last Report. It was detected 
in the Jamaica earthquakes tabulated by Maxwell Hall (Geop. Sup. M. N., 
vol. i. No. 2), and the period assigned as 21.001451m. But this led to the 
examination of a long series of Italian earthquakes (kindly lent by Mr. R. D. 
Oldham for the purpose), which very soon indicated a correction of one whole 
period in ten years. The discussion of the Italian series is not yet completed, 
but the Jamaica series was revised (Geop. Sup. M. N., vol. i. No. 3), for the 
new period 0,014584282d. or 21.001366m., which was found to suit them better. 
But there were fluctuations of six months’ period and of four years’ period in 
the position of maximum, the former with a range of 7.8m. (3.9m. on either 
side of the mean position), the latter with an even greater range of 14.0m. 
(7.0 on either side). A further fluctuation of fifteen months’ period was 
detected after the paper had been sent to press, but is discussed in a supple- 
mentary note to the paper. This fifteen months’ period had already been 
detected in ‘ Earthquake Phenomena’ (see the B.A. Report for 1912), and its 
elimination from the Jamaica series had the satisfactory result of removing a 
double maximum which previously affected the mean curve. 
We thus have three new periodicities in addition to the main one of 21m. : 
.viz. 4 years. 15 months, and 6 months. The 15 month (or 104/7 months more 
accurately) has independent support as above stated, though as yet we have 
no hint of its origin. The 6-month period, if real, is doubtless connected with 
the year. There remains the 4-year period which stood unsupported. But it 
was noticed that it affected the general frequency of Jamaica earthquakes 
(apart from its effect on the 21m. period); and also that of the Italian earth- 
quakes. An examination was made of the Chinese and Japanese long series 
and of other shorter series, and they were all found to be affected. The results 
are given in a paper now in the press (Geop. Sup. M. N., vol. i. No. 4). 
When the paper was read it had been inferred that the maximum travelled 
round the earth from east to west in 8 years (a double cycle, indicating a double 
polarity), but on checking the proof a serious error was detected in the Japanese 
reduction which rendered this view no longer tenable. It was found, however, 
that there was a travel in latitude from the Equator to the North Pole in the 
N. Hemisphere. What happens in the §. Hemisphere is donbtful, as we have 
only the single case of New Zealand to give information. But for what it is 
worth it indicates that the sweep is in the same direction (S. Pole to N. Pole). 
But for the further elucidation of these matters more material, and more 
accurate material, is required: and it would appear that our best line of advance 
at present is to continue the identification of evicentres and times for as many 
earthquakes as nossible. Hence this work of identification has been continually 
expanded (in the ‘ Bulletins’ for 1917 and their successors the ‘ Summaries’ 
for 1918) in spite of the consequent delay in catching up arrears. 
The General Propagation of Earthquake Waves. 
The announcement of the 21m. periodicity in the last Report had the incidental 
consequence of leading Dr. Jeans to undertake a new investigation of the whole 
auestion of earthquake wave propagation, and in a paper contributed to the 
Roval Society (Proc. R. S., A, vol. 102. 1923, n. 554) he points ont that in 
addition to the Rayleigh waves denoted by L. which travel with velocity 0.92 
(w/o), there are two whole series of surface waves of which the terminal 
members travel with velocities of VW(u/o) and W[(A+2u)/o]. Taking velocities 
sugeested bv the Oppav explosion. these waves would travel round the earth 
in 126m. and 222m., and Dr. Jeans suggests that returns of these waves to the 
epicentre (or perhaps the anticentre) may act as triggers for new earthauakes. 
‘Since this revort was sent to press a number of cases favourable to this view 
have heen noticed.] Independently of this possibility (which chiefly concerns 
periodicities) the paper must be of great value for the interpretation of seismo- 
crams, though no opportunity for testing it in this connection has yet been 
found. 
