19 (> 
DR. H. T. BARNES ON THE CAPACITY FOR HEAT OF WATER 
indicated by the bulb. I took special care to reduce this connection, which could 
easily amount to ‘01 or even more in the steam-jacket, to a negligible quantity, so 
that it could not have affected the F.I. as determined in my experiments to as much 
as ‘001°, and probably much less. Convection currents in the stem of the thermo¬ 
meter enclosed in the glass-tube were avoided in all my thermometers by lagging the 
leads down to within a few centims. of the bulb with cotton-wool. 
The insulation between the leads of the thermometers could be detected very 
quickly by a very simple adjustment in one of the box contacts, so that the battery 
in series with the galvanometer could be made to detect at once the smallest leak 
between the connecting and compensating leads. 
In correcting the differential readings to the air scale two of the ordinary parabolic 
formulae are combined. 
If pt x and t x be the platinum and air temperatures for one thermometer of the 
differential pair and pt 2 and t % be the same for the other, then for one we have 
and for the other 
h ~ Vh = 
8 
' tL _ A\ 
TOO 2 100/ 
U - pt % 
k\ 
100 / ’ 
from which we have, for the differential reading, 
(C ~ Ph) ~ (h ~ Ph) = bybbb ^ ~ + h ~ 100 )- 
The correction is always small, and amounts to T° in the extreme at the ends of 
the scale for a difference of 8°. It vanishes altogether at 50°, changing sign at that 
point. 
In determining the correction pt x and pt. 2 may be substituted in the right-hand 
side for a first approximation. A second approximation generally gives the correction 
with sufficient accuracy. 
Sec. 3 d. — Measurement of Time. 
The method of measuring the average rate of flow’ over the time of any experiment 
was to divide the total weight of water by the time of flow. This total time of flow 
was generally of the order of 15 minutes, or 900 seconds. The interval was recorded 
on a chronograph, which marked seconds 1 centim. long, from a standardized clock. 
The drum of the chronograph revolved at the rate of one revolution per minute, and 
the record of each second v r as made by a lateral kick in a continuous fine from a pen 
