ON EAETH TfiEMORS. 311 



double circular wall of the cellar, whicli is a circle of about 6 m. diameter, 

 the corresponding motion of the pendulum was still distinctly visible. 

 This effect of pressure formed a simple means of setting the pendulum 

 swinging, when it was necessary to determine the period of oscillation. 



It follows from these observations that the neighbourhood of the 

 column carrying a horizontal pendulum ought to be guarded against 

 receiving any additional weights during the course of observations. 



On the other hand, it was observed that the turning of the dome 

 which covers the east tower of the observatory had no disturbing 

 influence on the horizontal position of the pendulum, though the vibra- 

 tions of the soil could be distinctly felt by the observer. 



An experiment was also made to mitigate certain effects of micro- 

 seismic movements. In the first days of February 1889 a strong gale 

 was blowing, and the pendulum was found to be in a very disturbed state. 

 When trying to photograph the curve a broad and very irregular band 

 was obtained, which showed that the pendulum was not only constantly 

 swinging, but that it was subject to certain small changes of the vertical. 

 A point was fixed to the free end of the pendulum pointing downward 

 into a dish containing a mixture of water and glycerine. Although the 

 point only just touched the liquid, the pendulum was now unable to 

 swing, but the slow motions of its plane of equilibrium took pjace as 

 before. Photographs were taken for about a fortnight, and a very 

 remarkable case of earth pulsations, to which I will refer later, was 

 observed during this period. But, unfortunately, I was obliged to 

 abolish the glycerine, because it was evident that the position of the 

 pendulum was influenced by certain molecular effects. 



After these preliminary experiments, observations were taken between 

 April 1 and June 6, and again from June 18 until the end of September 

 1889. The column and pendulum were protected by a strong wooden 

 box nailed with tinfoil, which also covered the ditch and only contained a 

 small window, through which the light passed, and through which it was 

 possible to get at the foot-screws of the pendulum. The photographic 

 apparatus and lamp stood in a passage connecting the cellar with another 

 one lying to the north of the former, and were separated by a door, which 

 was also nailed with tinfoil and had two slits for the light to pass. This 

 door was only opened when it became necessary to bring the light point 

 back to the middle of the drum. On such occasions a thermometer 

 hanging near the column was read, and the same was done in the adjoin- 

 ing passage every other day when the paper was changed. During the 

 summer the temperature of the passage in which the lamp was burning 

 constantly was generally 1° or 2° C. higher than that of the cellar. The 

 door connecting the passage with the cellar to the north of it was only 

 opened twice every day, when the lamps or the paper had to be changed. 



At Wilhelmshaven the horizontal pendulum stood in a cellar of the 

 Imj)erial Naval Observatory, whilst the rest of the apparatus was placed 

 in an adjoining cellar lying to the south of the first. Openings were 

 made in the wall to let the light pass. All arrangements were very similar 

 to those described for Potsdam. Professor Boergen, the director of the 

 observatory, and his former assistant. Dr. Eschenhagen, very kindly 

 undertook to do all the work connected with the observations, and these 

 were carried on from March 7 until October 5. Amongst the several 

 difiiculties which at first presented themselves, and caused several inter- 

 ruptions of the observations, the most serious arose from the excessive 



