February 9, 1900.] 



SCIENCE. 



21T 



last will often quiver and rebound from the 

 impact of the wind, as if it had been hurled 

 against us in solid form. There is, of 

 course, some difficulty in obtaining a steady 

 illuminating flame during high winds, but 

 no disturbance of the mercury surface. 



The last previous observation of a star 

 had been made upon a circumpolar at upper 

 culmination, and nothing unusual was no- 

 ticed up to the time of leaving the tele- 

 scope. The last thread was taken nine 

 minutes before the oscillation was detected 

 at the Nadir, or at lOh. 41m. P. S. T. In 

 the interval the circles and micrometer 

 were read and the meteorological record 

 taken for refraction. The barometer was 

 again read, shortly following the cessation 

 of the disturbance, and no change had 

 taken place. The automatic recording in- 

 strument does not indicate any disturbance 

 in the pressure duriug this interval. 



The probable error of a telescope bisec- 

 tion is not greater than the hundredth part 

 of the oscillation, at the maximum ; so 

 that it is clear that nothing of that nature 

 was occurring at the time of observation. 

 Nor could it fail to be detected in the 

 transit record. 



The circular reading, itself, might not 

 indicate the beginning of the vibration ; 

 since it was of so slow and regular a char- 

 acter, that the telescope would probably 

 have not been disturbed, in its position 

 upon the pivots. An earthquake shock 

 usually changes the position of the tele- 

 scope in the wyes ; in one case a change of 

 3" was noted, as effect of a small, but per- 

 ceptible tremor. 



There is, thus, no way in which to fix 

 the commencement of the oscillation, be- 

 tween the limits of 10 h. 41 m. and 10 h. 

 50 m. The duratioa is extraordinary, es- 

 pecially considering its minute amount. It 

 was continuously observed for forty min- 

 utes. So impressive and startling was its 

 character, when noted in an instrument of 



such stable form of mounting, that the ob' 

 server had no hesitation in announcing the 

 coming of an earthquake for this locality. 

 Some previous experience served as the 

 basis for such a prediction. On May 7, 

 1894, a similar oscillation, but far less pro- 

 nounced in extent and duration, was de- 

 tected by the same means. It was followed, 

 after an interval of an hour and five min- 

 utes, by a slight earthquake, noted by three 

 observers on Mt. Hamilton.* 



In observing from two to three hundred 

 nadirs a year, these two instances are the 

 only ones noted of any vibration in the foun- 

 tain. The indications were that this pres- 

 aged a shock of a serious nature ; and the 

 chagrin, for an unfulfilled prophecy, is 

 easily more than counterbalanced by the 

 relief from the experience of a severe earth- 

 quake. 



The connection with the Mexican earth- 

 quake is pretty clear, taking into account 

 the slightly detailed accounts that have yet 

 reached us. That was most severe at 

 Colima, and Guadalajara, six or seven 

 persons having been killed, many injured, 

 and much damage to old and substantial 

 buildings having been done. The first ac- 

 counts of that disaster reached us on Mon- 

 day, through the daily papers ; and it ap- 

 pears that no further details are likely to 

 be printed in that form. At Colima, the 

 time was given at a quarter before twelve. 

 At Guadalajara it was recorded at mid- 

 night. These cities are 18°, or a little 

 more than an hour east of Mt. Hamilton : 

 so that the first statement of the time 

 would make the epoch near the beginning 

 of the vibration noted here. Until more ex- 

 act details of the times, recorded in Mexico, 

 are known here, it will not be possible to 

 check for closer agreement. 



The cities where the worst of the shock 



*Pub. Astr. Soc. Pao., 1894, Vol. VI., p. 184. 

 Catalogue of Earthquakes, E. S. Holden : Smithson- 

 ian Misoell. CoUec, No. 1087, p. 227. 



