328 EEPOKT— 1893. 



the period of which is nearly the same as the period of the ■waves on 

 January 5. 



Amongst the fifty-two days which show traces of this cuinous pheno- 

 menon during the time between 5 a.m. and noon, there are several on 

 which it appears developed into a large disturbance of the character of 

 an earthquake. Whether this is a chance coincidence, and earthquake 

 waves really happened to pass over Teneriffe so often during the same 

 time of the day, is a question which, of course, cannot be answered with 

 certainty. But it is much more probable to suppose that it was only 

 the intensity of the motion, in which these days differed from others, 

 like January 5 and April 7. Thus, if I bad possessed the means for ob- 

 serving real tilts, and not the swing of the pendulum on those days, 

 probably very large deflections would have been noted. 



It is very interesting to find that this phenomenon bears such a 

 decidedly local character at Orotava, for it never occurs at other times of 

 the day than in the early morning and forenoon. It generally begins 

 at a time when there is yet no trace of wind in the lower parts of the 

 island, though this circumstance may not appear of much importance, 

 because from many comparisons made it is nearly certain that the wind 

 exercises no influence whatever over the pendulum. The idea naturally 

 presents itself that the causes which produce the daily oscillation might 

 also be the reason why a certain period of the day is especially favourable 

 to the formation of earth pulsations. If the daily period be the effect of 

 a general motion of the ground, its condition of stress might vary, thus 

 affording the internal forces a better chance of producing such an effect 

 during one part of the day than during the other. 



I now turn to the description of some cases of earth pulsations lately 

 observed at Strassburg, but, as the observations are not yet finished, 

 I should like the following remarks to be considered as provisional only. 

 During the whole summer of 1892 photographs were taken, but never 

 did a trace of parth pulsations appear on them. The lines were not as 

 clear as I wished them to be, yet it was easily seen that in a general way 

 the pendulum was remarkably steady compared with what had been 

 noticed at the former stations. About the middle of October a slight 

 change was made in the lamp, which only produced a very insignificant 

 difference in the appearance of the curve. 



On October 19 earth pulsations suddenly appeared, such as had never 

 been seen before. The appearance of the curve is as follows vw^v^ and 

 is due to the imperfect figure of the light-point, which, instead of being 

 a small circle, has an oblong shape, with its axis a little inclined, thus 

 causing one branch of each wave to be more marked than the other. 

 This shows how important it is that the point of light should be as small 

 and as regular in shape as possible. During an interval of nearly ten 

 hours I counted about 200 successive waves, the duration of each being 

 nearly three minutes. At first the period of the waves was a little larger, 

 viz., 3 minutes 19 seconds, whilst towards the end it had gradually dimin- 

 ished to 2 minutes 43 seconds. The amplitude was a little less than tj'tt*!' 

 of a second, and remained constant throughout, thus giving the zigzag line 

 an extremely regular appearance. 



A very similar and yet more remarkable series of earth pulsations was 

 observed on December 22, when more than 300 waves were counted within 

 a period a little shorter than on October 19. Besides these two cases many 

 others occurred when the number of waves was smaller. The mean 



