48 
tlie time), we see that a rapid fall during the morning was 
checked soon after 2 p.m., a slight rise setting in, followed by a 
sudden jerk upwards, of 0'02 in., about 2-45. Then the fall 
Fig. 2. 
Midri Noon Midn Noon 
Fig. 3. 
9 Noon 3 6 B Midn. 3 . 6 
continued until 3-30, when there was a yet more marked 
recovery, followed by a further slight fall until 10 p.m. 
At midnight there began a rapid and steady rise of nearly an 
inch in 22 hours. 
The most notable incident, however, was overlooked by the 
engraver, and indeed had escaped my own attention until three 
