234 
DR. H. T. BARNES ON THE CAPACITY FOR HEAT OF WATER 
Sec. 6 .—Effect of Stream-line Motion on the Distribution of Heat in the 
Fine-flow Tube. 
* In Section 2 we discussed two conditions possibly existing in the fine-flow tube of 
the calorimeter, and explained generally the effect of these conditions on the 
temperature gradient of the glass surface and its influence on the heat-loss. In the 
present section I shall give briefly the experimental sequel to the theoretical 
considerations. In the light of the recent experiments of Professor Hele-Shaw,* 
so beautifully illustrating stream-line motion for water flowing at velocities under the 
critical velocity, and of Mr. T. E. Stanton! on 1 The Passage of Heat between 
Metal Surfaces and Liquids in Contact with them,’ the results might have been 
anticipated which I am about to describe. I do not think, however, that the effect 
of stream-line motion in fine tubes has been at all sufficiently appreciated. 
As the critical velocity at which the stream-line motion breaks down is so great 
(of the order of 10 feet per second) for tubes of from 2 to 3 millims. in diameter, the 
effect is inseparably connected with all experiments having to do with tubes of this 
size More especially does the flow tend to become linear, and to divide up into 
distinct and parallel lines, when a change of viscosity is introduced with a change of 
temperature. 
I must, in treating this part of the subject, apologise for the present incompleteness 
of my experiments, but I feel that I must give such as I have at present, not only to 
justify myself for the time and trouble I have taken to completely eliminate the 
effect of stream-line motion from my calorimetric measurements, but also as a 
beginning to some experiments on the distribution of heat from a metal conductor in 
water flowing at different velocities through fine tubes, which I hope to continue in 
the near future, and which I hope may at the same time throw some light on the 
difference in the rate of flow from the centre to the sides of the tube. 
My earlier experiments in 1898 were made with Calorimeter B, with a 2-millim. 
bore tube and central heating conductor, but with no special device for preventing 
stream-line motion.]; The measurement which we obtained of the mechanical 
equivalent at that time, as I have already pointed out, is affected to a certain extent 
by this, which was at once apparent when I came to use Calorimeter D, with a 
3-millim. bore tube. I undertook two sets of observations with Calorimeter D under 
two conditions, one with the heating wire, which was made of six strands of 6-millim. 
platinum wire, resting all along the edge of the tube, the other by drawing the wire, 
as best 1 could, straight through the centre of the tube. 
The results are as follows :— 
* ‘Proc. Inst. Naval Arcli.’ (1897), (1898); ‘Proc. Royal Inst/ (1899); ‘ Proc. Liverpool Eng. Soc.’, 
20, 37, (1898). 
f ‘ Phil. Trans.,’ A, vol. 190, 67 (1897). 
1 See note p. 219, supra. 
