115 
PROFESSOR HUGH L. CALLENDAR OX 
(33.) Methods of Eliminating Stream-Line Motion. 
It was evident from the equations above given that, since the conditions of linear 
flow gave rise to a systematic variation of the temperature of the flow-tube, which 
was directly proportional to the flow for a constant rise of temperature, it would be 
necessary to adopt some device for mixing the water in its passage through the tube 
so as to produce a nearly uniform distribution of temperature over the cross-section 
of the tube. The obvious method of securing this result was to employ a stranded 
conductor. This would diminish the superheating by increasing the available surface 
of the conductor, and would distribute the heat evenly over the cross-section of the 
tube, provided that- the strands were separated and arranged in such a manner as to 
break up the stream-lines. 
In order to verify the theory and observe the nature of the effects to be expected, 
I made some rough preliminary experiments on the superheating of various 
conductors in a 2 millim. tube with a steady flow of water. The general character of 
the flow, and the degree of mixing attained, were observed by the introduction of a 
colour-band of blue ink, after the method employed by Osborne Reynolds, and 
generally practised in hydraulic laboratories in studying the flow of liquids. The 
most instructive results were obtained with a stranded conductor consisting of 
5 strands of , 00G" pure platinum wire. My reason for selecting this particular size 
of wire was that I happened to possess a considerable quantity of it, and that its 
temperature-coefficient was accurately known, as it was regularly employed for 
maki ng thermometers. 
The resistance of the stranded platinum conductor, when carrying the heating 
current, was measured by the Wheatstone-bridge method, by comparison with a 
specially constructed platinoid resistance connected in series with it. The two were 
connected in parallel with a post-office box, by means of which the ratio of the 
resistances was observed. The resistance in the arm connected to the platinoid strip 
was 2000 ohms, the resistance in the adjustable arm corresponding to the platinum 
conductor was about 6000 ohms. As the platinoid strip remained practically 
constant, the resistance of the platinum could be taken as proportional to the 
resistance in the adjustable arm. The watts on the conductor were observed by 
means of a Weston ammeter and voltmeter of suitable ranges. 
The platinoid strip resistance, which was subsequently utilised for regulating the 
current, is shown very clearly at L in the bird’s-eye view (Barnes, fig. 15, p. 213). It 
consisted of a number of strips of platinoid about 1 foot long, one-half inch broad, and 
-fs inch thick, having a resistance of nearly yjjth of an ohm each. The ends of the 
strips were bent at right angles and amalgamated. They could be connected in series 
or parallel in a great number of different combinations by means of the mercury cups 
and copper connectors shown in the plate. For this particular experiment they were 
