OX IMPACT WITH A LIQUID SURFACK 17'. 



m-arer the sphere. Fig. 5 shows that the division of the air column is accomj>anied 

 by the formation of an annular vortex. Fig. 6 [7, 8, and 9] show the filling up of the 

 i-r.it IT. The vertical lines of bubbles recover their original positions ; but at the top, 

 where the lateral displacement has been greatest, it will be observed, especially in 

 tig. G, that the stream of bubbles turns over inwards, and it is on this that we rely as 

 proving that in part, at any rate, the hollow fills up by the pouring in of liquid 

 down the sides. There are unmistakable indications of this same curling over in the 

 negjitive of fig. [8] when the jet has begun; but the stream of bubbles is so faintly 

 visible that it would disappear in any reproduction. 



The next step in the investigation was to examine in more minute detail the splash 

 of a smooth sphere. For this purpose we procured a number of very hard highly 

 polished steel 1 tearing Kills of three sizes,* viz. : 



(1) )} inch (= 19'0 millims.) in diameter. 



(2)f (=15-9 " ) 

 (3) it .. (= 127 ) 



Several of each size were subsequently coated with an electrolytic deposit of nickel, 

 \\lik-h took a still higher polish. 



In the year 1897 several hundred splashes were observed with these spheres. 

 Previous observations, as recorded in Series IV., V., and VI. of the earlier paper, had 

 shown that when a well-polished sphere enters with a not very great velocity, the 

 liquid rises over it in a thin sheath, covering it completely before it is entirely below 

 the level of the surface, and that such a sphere consequently takes down no air what- 

 ever into the liquid ; also that the motion of the liquid is different, from the earliest 

 moments of contact, from that set up by the entry of a rough sphere. We had, 

 however, noticed that with a much-increased height of fall the smooth sphere took 

 down air and behaved like a rough one. 



It appeared possible to us that there was some critical velocity which might mark 

 the change very sharply, and accordingly our first observations were directed to the 

 examination of this point. It was not necessary for this to use any instantaneous 

 illumination; the magnetically-controlled "catapult" release was still employed for 

 letting the spheres fall, but the observations were conducted in the full light of the 

 laboratory. The results of some hundreds of observations may be summarised as 

 follows : 



With any one sphere always carefully polished in the same way just before letting 

 it fall, a height of fall is soon reached at which the splash ceases to be " completely 

 airless" or smooth, but is followed by the rise of bubbles, at first a shower of very 



* \Ve afterwards, as will 1m seen, employed spheres, 25-7 millims. in diameter, of polished serpentine, 

 ami may here state at mire that we have not found that the size of the sphere 1ms, within the limits 

 mentioned, any noticeable iiiHueni-e on the course of the splash. 



" A '"' 



