﻿1010 Research Staff of the G-. E. C., London, on 



(c) Drawing Device. — The fine wires (tungsten or constan- 

 tan) used as cylinders were cleaned superficially by heating 

 in a reducing atmosphere. They were then wound on a 

 bobbin P (see fig. 2) above the surface of the liquid, and 

 drawn thence round an idle pulley L below the surface of 

 the liquid and finally round the winding bobbin N 7 which 

 was rotated at a regulated speed. As an approximation to an 

 infinite slab a copper strip about 1 inch wide and 0*05 cm. 

 thick was used. It was thought that if such strip were 

 drawn through the liquid from a bobbin outside it, the liquid 

 might be cooled appreciably when the drawing was rapid. 

 Accordingly the strip was originally wound on a bobbin, 

 wholly immersed in the liquid, and drawn thence direct to 

 the winding bobbin. 



(d) Estimation of thickness. — The thickness ot the layer 

 on the fine wires was determined by weighing a known 

 length of the coated wires on a torsion balance (designed in 

 these Laboratories) capable of estimating a few milligrams 

 with an error of 0*0001 milligram. The coating was then 

 dissolved off and the bare wire weighed. The layer on the 

 strip is uniform only at some distance from the edges ; 

 accordingly a known area was cut from the central portion 

 and weighed before and after removal of the layer. The 

 density of the solid layer was determined on a sample of 

 convenient size. 



(e) Liquids used for coating. — Molten waxes were first 

 used as the liquid for coating the solid surfaces. It was 

 found that filtered beeswax and carnauba wax were most 

 suitable. 



A range of viscosity from 1 to 100 centipoises was 

 obtainable within a temperature range from 50° to 140° C. 

 In some experiments the viscosity was increased by adding 

 fine insoluble powders to form suspensions. Waxes were 

 expected to be particularly advantageous, because their 

 surface tension is low and therefore not likely to be changed 

 by grease and other impurities of low surface tension. But 

 it was found in the course of the work that aqueous solutions 

 could be used with convenience ; for by the process of drawing- 

 surface impurities were removed very rapidly, so that after a 

 short length had been drawn the surface tension was con- 

 stant and normal. Some of the observations recorded were 

 made with aqueous solutions of sodium silicate, which were 

 dried on the wire by passing it through a small electric 

 furnace a short distance above the liquid. 



