178 



MESSRS. A. M. NVOKTHIMiToN \\1> FJ. S. 



way for the emergent jet (cf. Series II., figs. IG, 17, ;m<l I \. :nil Series III., tig. 18, 

 of the earlier paper). 



In order to obtain, if possible, some LU&cmatton ;is t<> the actual motioo of tin- 

 liquid, a variety of experiments was mode with coloured binds and floating b<xlics, 

 ;ind finally we bit upon the methcxl of letting the sphere fall into dilute sulphuric 

 acid between two electrodes (the bared extremities of two insulated vertical copper 

 wires), from each of which there ascended slowly a vertical stream of very minute 

 bubbles liberated by electrolysis. The velocity of ascent of these bubbles, though it 

 was not always quite slow enough to avoid eddying motion, was so slow (about l\ 

 centims. per second) that the gravitative displacement of any one bubble during ;i 

 splash is practically negligible, and the line of bubbles may be therefore taken as a 

 line of marked liquid whose displacement can thus l>e studied, for the bubbles are so 

 small that it may, we think, be safely assumed that they did not in any way interfere 

 with the motion of the neighbouring liquid. On account of this minuteness the 

 stream of bubbles is not always easy to see in the photographs, and in printing from 

 the negatives long exposure in bright sunshine is necessary. 



It will be observed that whereas in fig. 1 of Series XII. (here reproduced) each 

 stream is delivered from the top of the electrode ; in figs. 2 and 3 the stream has 

 been swept off the electrode, especially on the left-hand side where the electrode is 



Fig. 1. 



Series XII. 

 Fig. >. 



Fig. 3. 



