﻿1012 Research Staff of the Gr. E. C, London, on 



Moreover, it made no difference to the thickness of the 

 coating whether the wire was drawn vertically from the 

 liquid or horizontally through the top of the surface formed 

 by a rotating wheel within the liquid. Surface tension 

 must therefore be the dominant force. This result is in 



accordance with eqn. (9), for the least value of — — 2 in the 

 observations was 300. P^ v ° 



However, the possible effect of instability due to surface 

 tension must be considered. It was clear immediately that 

 the layer originally formed was unstable, for the solidified 

 coating on the wire was not even but was broken up into 

 blobs. These blobs were beautifully symmetrical and evenly 

 distributed when viewed under a microscope. Photomicro- 

 graphs (1), (2), (3) are shown in fig. 5 (PI. VIII.) . But 

 formation of the blobs evidently occurs after the wire has left 

 the liquid surface ; for it does not depend on the velocity of 

 drawing or the viscosity within fairly wide limits. No. (1) 

 was obtained at one-half the speed of No. (2) and at twice 

 the viscosity ; in these two cases there is approximately the 

 same thickness of coating. No. 3 was obtained at a higher 

 speed than No. 2, but at the same viscosity. Here the 

 coating is obviously thicker. 



That the blobbing took place after the liquid coating had 

 been withdrawn from the liquid was further demonstrated 

 by drawing a wire through glycerine ; a very thick film 

 was obtained which remained smooth and even for some 

 inches above the surface of the liquid before it could be seen 

 to break up. 



A systematic study of the film thickness as a function of 

 viscosity and surface tension and velocity of drawing as well 

 as the radius of the wire was undertaken. 



The following were the limits of the variables : — 



Radius of wire r = 0'00075-0"01 cm. 

 Velocity of drawing v = 5'5-66'0 cm. /sec. 

 Viscosity 77 = 1-100 centipoises. 

 Surface tension 7=36-62 dyne/cm. 

 Density /) = 0'8-l*4 grm /cm 3 . 



The limits of v and rj that could be investigated were set, 

 on the one hand by the impossibility of measuring very thin 

 layers, on the other by the thickness of the film ; if t were 

 greater than r the blobs which formed fell off the wire. 



The results are shown in fig. 6 by plotting ( — ) against 



