ON SEISMOLOGICAL INVESTIGATION. 105 



No. 371.— V. S=60. 



III. Earthquakes and Timekeepers at Observatories. 



That earthquakes we can feel frequently accelerate, retard, or stop 

 clocks with pendulums is a fact well known, but the extent to which 

 cryptoseismic disturbances which sweep over the whole surface of our 

 globe many times per year aflfect this class of timekeepers has not yet 

 been investigated. 



Father J. de Moidrey, S.J., of the observatory at Zikawei, gives me 

 the following notes on this subject. On June 12, 1897, ' an excellent clock 

 facing north lost 4m. 44-5s. in the afternoon, whilst another, almost 

 identical, fixed to the same brick pillar, but facing east, was undisturbed 

 (rate 0-ls.). Secchi's barograph shows a slight stroke at llh. 25m. 

 G.M.T., corresponding to an oscillation of 1 mm. of the quicksilver. 



'A fast moving barograph (mercury) shows a spot at llh. 2.3m., indi- 

 cating a swing of the mercury of 0-2-5 mm. This increased to 0'50mm. 

 and died out suddenly. 



' The magnetographs, declinometer, bifilar and Lloyd's balance were 

 all disturbed, although it was a day of perfect magnetic calm.' 



On this day, at llh. 5m. G.M.T., a violent earthquake took place in 

 Assam. The large waves of this would reach Zikawei at llh. 21m. G.M.T., 

 or 7h. 26m. 43s. p.m. local time. 



In a second letter Father Moidrey writes : 



' On June 4, 1898, about midnight, our north clock lost about four 

 seconds. That same night at a watchmaker's in Shanghai several clocks 

 (six, I believe), all facing north or south, were stopped. Nothing else was 

 noticed by the watchmaker, M. Vrard, who in his surprise telephoned to 

 the observatory to ask what was the matter. Nobody in the town felt an 

 earthquake, nor was one referred to in the newspapers. A missionary 

 at Nankin had his clock stopped the same night, but did not notice any 

 other phenomena. Our magnetograph and thermograph recorded a shock 

 at 16h. 24m. 17s., June 3, G.M.T. On that day there was an earthquake 

 at Chemulpo, Corea.' 



We' are here evidently dealing with an earthquake recorded on June 3 

 at 17h. 14m. at Shide, and also recorded at Kew, Nicolaiew, and Potsdam. 



From the ' Bulletin Mensuel ' of Zikawei, third quarter, 1897, we learn 

 that in the night of September 2 the two clocks were stopped and the 

 magnetographs were disturbed at 1.42 (September 1, 17h. 36m. G.M.T.). 

 Nothing was felt. This may refer to an earthquake recorded at Shide, 

 September 1, 18h. 29m. G.M.T. 



Although Professor E. C. Pickering writes me that on September 3, 

 10 and 23, 1898, which are dates for heavy earthquakes in Alaska, and 

 on September 20, when there was a severe earthquake in Asia Minor, 

 there were no noticeable changes in the rates of the clocks at Harvard 



