216 REPORTS ON THE STATE OP SCIENCE, ETC. — 1920. 



installation of electric light in the laboratory. This has much facilitated the 

 operations of changing fiiins, comparing clocks, &c., but the gas-jet is retained 

 for the photography. The room has further been cleaned and whitewashed, ana 

 an outer door has been added shutting it off from draughts. It is now a very 

 convenient laboratory, and is large enough for the erection of at least one 

 more machine, when one is available. 



The Milne-Shaw machine formerly erected at Eskdalemuir for direct com- 

 parison with Galitzin records has been now transferred (on loan) to the Royal 

 Observatory, Edinburgh, and readings have been received from July 4, 1919. 

 The situation seems peculiarly liable to microseismic disturbance, obviously 

 connected with wind. 



The instrument mounted in the ' dug-out ' near West Bromwich has given 

 •some interesting results as regards these microseisms on which Mr. Shaw writes 

 a special note at the end of this report. 



Various other instruments are being constructed as rapidly as present difficul- 

 ties permit. 



Milne-Shaw machines have recently been dispatched to Cape Town, Montreal, 

 Honolulu, and Aberdeen. Others are being made for India, China, Egypt, New 

 Zealand, Canada, and Ireland. 



Bulletins and Tables. 



' The Large Earthquakes of 1916 ' have been collated and published as a 

 single pamphlet of 116 pages, but there are great difficulties in obtaining satis- 

 factory determinations of epicentres for the later war years, which have delayed 

 f ui-ther publication. 



The corrections to adopted tables have not yet been completed. 



Earthquake Periodicity. 



The study of long periods in the ' Chinese Earthquakes ' directed attention 

 to a period near 260 years. This was in the first instance identified as 240 years 

 (' Mon. Not. R.A.S.,' Ixxix., p. 531) as mentioned in the last report, and Mr. De 

 Lury pointed out that this value also suited tree-records (Pub. Amer. Ast. 

 Soc. 1919). But an investigation on the secular accelei-ation ol the Moon by Dr. 

 Fotheringham recalled attention to a value nearer 260 years, which was also 

 found to suit the tree-records (' Mon. Not. R.A.S.,' Ixxx., p. 578) over the same 

 period. Ultimately a much longer series of tree-records was obtained (Mr. A. E, 

 Douglass's compilation from 1180 B.C.) and a full analysis of these, now in the 

 press ('Mon. Not. R.A.S.,' 1920 Supp. No.), suggests a double periodicity, with 

 components of approximate lengths 284 and 303 years. Long as it is, the series 

 of tree-records is not long enough to separate these components themselves : the 

 evidence for separation is provided by the harmonics, especially the third 

 harmonic, which shows components of 101 years and 94'4 years clearly separated, 

 the former and longer being the stronger, whereas in the main terms the shorter 

 period is the stronger. The second harmonic of the longer period, i.e., half 303, 

 or, say, 152 years, is) quite possibly the 156-year period referred to in the last 

 report. 



These results have been obtained so recently that their full relation to the 

 earthquake records have not yet been worked out. But a welcome confirmation 

 may be mentioned. In the 'Bull. Seism. Soc. of America,' vol ii.. No. 1, Miss 

 Bellamy found a later list of ' Chinese Earthquakes ' compiled by N. F. Drake. 

 It is not entirely independent of the catalogue already studied (compiled by 

 Shinobu Hirota in 1908 and mentioned by Drake as having been received too 

 late for inclusion or comparison), but it differs from it in one important respect, 

 being copious in the later centuries where Hirota 's catalogue is scanty. Further, 

 it is confined to ' destructive or nearly destructive' earthquakes, so that the 

 records are probably more pi'ecisely comparable inter se, although they still 

 Bhow a large increase about a.d. 1300, which must be attributed to greater 



