302 REPOKT— 1893. 



heavy mass employed is about 84 lb., and consists of a part of the 

 cast-iron pipe of which the foundation is made. When it was placed 

 at successive distances of a toot, from 2 to 9 feet, nearly west of the 

 pendulum, the tilting of the column was inappreciable ; and this was also 

 ihe case when it was placed alternately at distances of 2 and 9 feet. 

 If, however, it was put alternately nearly west and east of the pendulum, 

 and close to the wooden cover of the drain-pipe, the tilting, though very 

 slight, was perceptible, and apparently in the direction away from the 

 heavy mass. For the reasons above stated, however, the results were 

 not altogether concordant, so that these experiments prove little more 

 than the nearly complete isolation of the stand. 



Earth Pulsations of June 3 and 6, 1893. — At 5.4.3 p.m. on June 3 the 

 image was, as usual, perfectly still, and was adjusted to the cross-wire 

 without difficulty. At 6.29, when I went to take the next reading, it was 

 marching slowly and steadily from side to side of the field of view. At 

 first I timed each separate oscillation, and found their period to be nearly 

 regular, and between 15 and 17 seconds in length. At this time the 

 range of movement was greatest, but it was impossible to determine its 

 amount exactly, owing to the shifting of the mean position of the image 

 produced by the heat of the gas-jet. The movement continued to be 

 considerable for about five minutes, then it gradually diminished, and by 

 6.37 was decidedly less, perhaps half that at 6.30. Between 6h. .35m. 15s. 

 and 6h. 37m. 40s. the image crossed the wire fourteen times, the average 

 period of the seven complete oscillations being therefore 20" 7 seconds. 

 At 6.42 the image had become steady again. I then tried to measure the 

 range of the movement when it was greatest from my recollection of the 

 average limiting positions of the image, and found the scale-value to be 

 0'44 inch, and the corresponding angular value about ^th of a second. 



At 6.46 the oscillations began again, and continued from this time 

 until 8.13, with a few short intervals, when they either ceased or were 

 almost imperceptible; but throughout this time the range was always 

 very small, never, I believe, exceeding a^^th of a second. The oscillations 

 were also more constant in range and in their mean position, so that the 

 range could be determined by making the image in each limiting position 

 coincide with the cross-wire. Between 6h. 48m. 40s. and 6h. 53m. 40s. 

 there were sixteen complete oscillations, giving an average period for 

 -each of 18'8 seconds. At 6. 56 the range was about -,}^Ji\l of a second, and 

 at 7.8 about ^'oth of a second. After this the oscillations became smaller, 

 and it could just be seen that the image was in motion. This continued 

 until just before 7.19, when the movement became greater again and 

 more regular. Between 7h. 19m. Os. and 7h. 25m. 43s. there were twenty 

 complete oscillations, the average period being 20'2 seconds ; and the 

 range immediately after the latter time was about ^Tyth of a second. Then 

 the oscillations again decreased, and at 7.36 the image was nearly steady, 

 occasionally being just visible on one side or other of the cross-wire. 

 Between 7h. 42m. 58s. and 7h. 45m. 50s. there wei'e five complete oscilla- 

 tions, but the movement was so slight that one or two of them may have 

 been missed. At 7'54 I left the cellar, returning about ten minutes later. 

 About 8.8 the last series of oscillations began. Between 8h. 8m. 15s. and 

 8h. 11m. 20s. ten complete oscillations were counted, the average period 

 being 18*5 seconds, and the range immediately afterwards was found to 

 be about yio^^ o^ ^ second. After 8.13, though I watched almost con- 



