238 
DR. H. T. BARNES ON THE CAPACITY FOR HEAT OF WATER 
these values as being true for any rise where the temperature distribution is uniform 
throughout the water column and stream-line motion avoided, then the value of d 
may be calculated. The two values of the heat-loss per degree rise so obtained were 
for a flow of 
'54000 gramme per sec., '04445 watt. '27300 gramme per sec., '04965 watt. 
The latter value has not to he corrected for the small conduction effect for the small 
flow on account of the method of treatment. Hence the value of d comes out 
— '00464, from which 
J = 4'2 (1 - 00464) = 4*1805 joules, at 30° C. 
The value of J at the same temperature, obtained with the other calorimeters for anv 
rise of temperature when using the various devices for obviating stream-line motion, 
is 4*1780 joules, which agrees to 1 part in 2000 with this value. This is quite as 
good an agreement as can be expected from the manner of treating the observations, 
and the want of agreement in the observations themselves for the small flow, which 
is no doubt occasioned by the uncertainty introduced by the stream-lines. 
Sec. 8 .—Experiments between 0° and 100° C. 
As soon as it became clear that the main cause of error had been removed in 
eliminating the effect of stream-line motion in the calorimeter, I commenced a series 
of experiments to extend over the entire range of temperature. These experiments 
are summarized in the following table from I. to LV., and include upwards of 46 
