BETWEEN THE FREEZING AND BOILING-POINTS. 
201 
Table XYII. — Summary of Clock Comparisons. 
Month. 
Greatest loss 
in seconds. 
Least loss 
in seconds. 
Mean loss 
in seconds. 
Mean loss 
corrected for rate 
of chronometer. 
May, 1898 . . . 
26 
25 
26 
19 
June, ,, ... 
33 
27 
30 
23 
July, „ . . . 
36 
32 
34 
27 
August, „ . . . 
37 
33 
35 
28 
September ,, ... 
39 
36 
38 
31 
December, ,, ... 
29 
23 
26 
19 
January, 1899 . . . 
25 
21 
23 
16 
February, ,, ... 
22 
18 
20 
13 
March, ,, ... 
19 
18 
19 
12 
April, ,, ... 
20 
18 
19 
12 
May, „ . . . 
23 
21 
22 
15 
June, ,, . . . 
35 
29 
32 
25 
September, ,, ... 
38 
37 
38 
31 
October, ,, . . . 
40 
38 
39 
32 
November, ,, ... 
39 
31 
35 
28 
February, 1900 . . . 
23 
22 
23 
16 
March, ,, ... 
23 
19 
21 
14 
The rate of loss diminishes in winter and just doubles during summer. This is 
probably due to the effect of the dry furnace heat in the building during winter on 
the wooden pendulum of the clock, in contrast to the more humid climate of the 
summer months. The furnace fires are started about the month of November and 
discontinue some time in April. 
Sec. 3e ,—Measurement of Weight. 
In all the older methods of calorimetry, the question of evaporation of the water 
becomes a serious one for consideration. In the present experiments there were no 
such difficulties. The stream of water flowing from either one of the two nozzles on 
the outflow end of the calorimeter was caught in a weighed flask, which was fitted 
with a rubber stopper through which the nozzle passed. Through a second hole in the 
stopper a tube was fastened containing calcium chloride, so that the air in the flask, 
displaced by the inflowing water, passed through the calcium chloride. At no point 
between the calorimeter and the interior of the flask did the water come in contact 
with air. 
In fig. 8 a drawing of the flask is given, showing the position of the calcium-chloride 
tube. The hole through which the nozzle of the calorimeter is thrust is closed, when 
the flask is removed, by a glass stopper. A similar stopper closes the end of the 
calcium-chloride tube and prevents the absorption of water vapour from the air. 
The capacity of each of the two flasks, which were used in the experiments, was 
about 750 cub. centims., but the amount of water weighed in them was never more 
VOL. CXOIX.—A. 2 D 
