156 



REPORT — 1894. 



experience any changes of temperature throughout the day. On the 

 enclosed copy the diurnal oscillations are shown very distinctly. 



In my letter of August ^f, 1893, I pointed out the three principal 

 kinds of movement which the horizontal pendulum experiences : (1) the 



annual or long-period deviations 



Fig. 4. — Earthquake Disturbances of 

 March 21-22, 1894. 



which M. de Rebeur-Paschwitz 

 calls ' Nullpunctbewegungen ; ' (2) 

 the diurnal deviations ; and (3) 

 seismic disturbances. To these must 

 be added (4) disturbances during 

 storms, probably arising from the 

 movement of the building, produc- 

 ing tremors in the ground ; and 

 (5) periodic deviations of short 

 period, in all probability of seismic 

 origin. 



The deviations (1) are shown 

 in the continual movement of the 

 pendulum in the same direction,, 

 with slight digressions, lasting 

 several months, apparently during 

 the transition from winter to sum- 

 mer, and vice versd. In the present 

 position of my instrument the pen- 

 dulum inclines towards the south in 

 spring and summer, and towards 

 the north in winter. . . . Here I 

 must remark that the amplitude of 

 the annual changes of temperature 

 in the cellar where the instrument 

 is placed does not exceed 6° R. 

 (13°-5 F.), whilst the changes of the 

 diurnal period are quite insensible. 

 In fig. 4 are seen two seismic 

 disturbances on March 21-22, 1894. 

 The second of these (beginning at 

 Oh. 43m. Nicolaiew mean time)' 

 coincides with the disturbance ob- 

 served at all the Italian seismic 

 stations, and also registered by the 

 magnetographs of Pola, Potsdam, 

 and Wilhelmshaven (see * Boll. 

 Meteorico dell' XJfficio centrale . . . 

 al CoUegio Romano,' No. 135, Sup- 

 plement© 103),andwas probably pro- 

 duced by the earthquake in Japan 

 (7h. 27m. 49s. Tokio mean time). 

 Four to five hours earlier another 

 rather strong disturbance is seen 

 on the photogram, which does not 

 coincide with any observed earth- 

 quake, but which was also registered by the horizontal pendulum at 

 Charkow 



