ON EARTH TREMORS. 



323 



to add that in the present investigation no account has been taken of a 

 purely local effect of temperature which is caused by a difference in th« 

 length of the two foot-screws east and west of the pendulum. If they 

 are not equally long, a change of temperature will produce a small tilt of 

 the instrument. This, however, would only be noticeable in the motion 

 of the zero-point, for the daily change of temperature in the cellars was 

 practically zero. But also in the former it would only account for a small 

 motion, such as, for instance, was found for Orotava. Considering the 

 dimensions of the instruments, I find that a difference of temperature of 

 1° C. produces a tilt of 0"'0164 multiplied by the difference in the length 

 of the screws expressed in millimetres. The surface of the column 

 which carries the pendulum ought therefore to be made as neai'ly hori- 

 zontal as possible, supposing the instrument to possess an entirely sym- 

 metrical form. 



(6) Besides those changes which we have until now considered, and 

 which by the nature of their causes can never surpass certain limits, the 

 zero-point is subject to others of particular interest, which are probably 

 due to geological causes, viz., the slow folding of the earth's crust. By 

 the aid of the diagrams drawn I have tried to fill out the gaps in the 

 observations, and to construct curves which represent approximately the 

 motion of the zero- point, corrected for temperature and barometric pres- 

 sure, during the whole period of observation. Starting from an arbitrary 

 zero, I have obtained the following values, in which an increase indicates 

 a motion towards the east, corresponding to a tilt in the same direction : — 



In compatiiig the values for Wilhelmshaven I supposed that during the 

 month of June there was no considerable alteration in the position of the 

 zero- point as it is indicated by the form of the diagrams before and after 

 the interruption. The observations at Orotava are complete, whilst at 

 Potsdam the interruption is so short that the given values may be con- 

 sidered as very nearly accurate. 



Thus we find that the pendulum showed the greatest motion at Pots- 

 dam, where its installation was certainly the most favourable of all. 

 Observations there were commenced about five months after the column 

 was completed. During the whole of April and the first part of May the 

 column inclined towards the west; it then turned and a tilt of Il"'2 

 towards the east began, which lasted until the middle of June, when again 

 it inclined towards the west, moving through an angle of 14" until the 

 end of the observations. 



Y 2 



