ON IMI'ACI WITH A LlgllD 



177 



same dimensions in air. Thus the cylinder finally divides into two portions, one <>(' 

 \\liicli follows in the wake of the sphere as an attached bubble (figs. 16 and 17), 

 \\hile the other rapidly fills up, apparently, or at any rate in part, by the pouring in 

 <>!' liquid round the rim of the liasin. The convergence of this inward flow corre- 

 sjxmds to an increasing velocity as the axis is approached, and results in the very 

 rapid upward spirt of the jet that is so well shown in Series VIII. of the previous 

 paper. 



Our opinion that the basin does fill up by a motion of this kind is based upon the 

 evidence of the photographs of Series XII., which will be explained shortly. Mean- 

 \\liile it is convenient to point out that Series XI. (here shown by drawings) gives the 

 under-water phenomena corresjx>nding to Series IX. of the previous paper. In this 

 the sphere (still rough) is let fall from a greater height, 50 centims., and the 

 crater thrown up closes and forms :i bubble which subsequently opens again and makes 



Fig. 1. 



Series XI. 



Fig. -1. 



Fig. 3. 



Fij;. I. 



Fi*. :.. 



\l.. OXCIV. A. 



2 A 



