ON IMPACT WITH A LIQUID SURFACE. 
143 
scale |- of actual size (linear). The first 7 or 8 show the evolution and rise of the 
crater to its maximum height, which is attained in about two hundredths of a second ; 
this crater then remains poised with but little change for another hundredth of a 
second (figs. 8 to 12), and then (figs. 13 to 20) in about two and a half hundredths 
more widens out and subsides till nothing but a lobed rim is left above the surface 
surrounding a central hollow (fig, 20). This is followed by the rise of a central 
column carrying the liquid of the original drop on its summit. The distinction 
between the more transparent water at the top and the comparatively opaque adherent 
milk below is quite observable in the original of Photograph (22) though hardly 
apparent in the reproduction, and, in all, the lamp-black carried down by the drop 
is seen to be collected chiefly at the summit. The rise of the column takes 
about second (figs. 21 to 26), and its subsequent subsidence (figs. 27 to 29) 
about second more. In Photograph 24 is seen the first appearance of an out¬ 
ward-spreading ripple. Photographs 26 to 30 show how the base of the column 
gi'adually flattens down into a “ cake” of liquid, whose edge marks the position of the 
next well-marked ripple, while figs. 3], 32, and 33 show how by the oscillations of the 
centre, a third “ cake ” is superposed on this, contributing the third outward-spreading 
ripple and so on. We were not able conveniently to follow the phenomenon further, 
through the laboratory being too low for the height of fall necessary for the timing- 
sphere at these late stages. 
It should be mentioned that it is known (Worthington, ‘ Proc. Roy. Soc.,’ 1882, 
vol. 34, p. 217) that the subsidence of the central column gives rise to a vortex ring 
that descends through the liquid."^ 
The reason for mixing milk with the water into which the drop fell was to secure 
something which would photograph. It was found that the addition of milk in the 
proportion of about one part of milk to three of water, though it must have reduced 
the value of the surface tension, did not make any decided or very noticeable change 
in the phenomena. If pure milk was used the crater thrown up was indeed some¬ 
what higher and had longer arms, indicating a smaller efficiency of the surface tension 
in opposing the rise of the liquid. 
Series II. (Plate 2). — This gives, in 37 photographs, the spla.sh of a milk drop of 
diameter '75 centim. [circa) falling 100 centims. into water. Scale f linear, as far as 
No. 18 (the single quartz lens being used). Thence onward f linear (with the ordinary 
lens of the camera). In Nos. 15-18 a little milk was added to the water to make the 
photographs clearer, and from 19 onward a good deal of milk (about one part of milk 
to three of water). In this splash the crater rises to a greater height and closes 
completely over the central hollow, opening again, however, very shortly afterwards 
to make way for the column that rises from the base, and whose subsidence produces 
“cakes” as before. In Photograph 18 the bubble has apparently not yet burst 
before its top is struck by the column rising inside. Sub-group A consists of special 
* See also Thomson and Newall, ‘ Proc. Roy. Soc.,’ 1885, vol. 39, p. 417, 
