ON IMPACT WITH A LIQUID SURFACE. 
145 
it will contract under the influence of the surface tension exceeds the inward 
acceleration of the cylindrical walls of the crater arising from the same cause, so 
long as the diameter of the rim is less than 1-61 of the thickness of the wall. Although 
this must be a vera causa in determining the more rapid contraction of the uj^per 
portion of the crater, 3 mt it may not be the sole cause. Photograph 8, and particularly 
Photograph 2^, of the sub-group suggest that there may be a diminution of air- 
pressure within the crater, owing to the descent of its base whereby the crater is in 
part forced in by excess of external air-pressure. (See also, with a lens, Photograph 5 
of Series IX.) 
It appears to be characteristic of all closing bubbles that the arms are inclined 
outwards (see also Series II. and IX.), as if they were being dragged in by the 
contracting rim from which they spring. In an opening bubble they are much more 
erect (Series III., 13 and 14), or even inclined inwards (No. 18). A late stage 
in a bubble is also differentiated from an early one by the greater smoothness of its 
surface, and by the absence of very small drops in air above it, such smaller drops 
having apparently had time to agglomerate into larger. 
Splashes of Solid Sp)heres .—The remaining Series exhibit the splash of solid 
spheres. It was already known (Worthington, ‘Proc. Poy. Soc.,’ 1882, loc. cit.) 
that the disturbance set up by a very smooth and well-polished sphere is quite 
different from that due to the impact of the same sphere when rough or wet, and it is 
a matter of great interest to find that the difference is quite pronounced from the first 
instant of contact. 
Series IV. (Plate 4) gives, in 17 photographs, the splash of a well-polished sjihere 
of ivory, 1'9 centim. in diameter, falling 60 centims. into water mixed with milk, con¬ 
tained in a glass bowl about 1 foot deep and 9 inches in diameter (scale f linear). In 
order to secure that the splash shall follow the lines here recorded, the polishing with 
a dry cloth or wash-leather must he repeated just before each observation, and after 
this the sphere must be handled as little as j)ossible ; with these precautions a stone 
sphere behaves in just the same way. Photographs No. 2 and No. S'* show that the 
liquid rises over and surrounds the sphere with a thin close-fitting sheath. Figs. 5 
and 6 show a subordinate side-sheath, which, without doubt, wms due to the fact that 
the sphere had a crack in it, which occasionally carried down air with it, and 
disturbed the symmetry of the splash. In Photograph No. 7, however, the symmetry 
is complete, and it is nearly so in Photograph No. 8. When any failure of the 
polishing occurs, the liquid is kept away from the sphere, and the splash, instead of 
being almost noiseless, is accompanied by a sound of bubbles rising to the surface and 
bursting. Photograph No. 4 shows this driving away of the liquid, and our note-book 
records that this splash was attended by “ noise, bubbles,” &c. In Photograph No. 8 
there is a similar sign of roughness on the right-hand side. Very important is the 
information given by the shadow thrown across the surface in such figures as 8 and 9. 
'* The light mai’kiug on the right .side of Photograph No. .3 is due to a flaw in the negative. 
MDCCCXCVII.—A. U 
