2 we have an annular rim, which in Figs. 13
and 14 is seen to topple into lobes by which the rays are united in
pairs, and even the special rays that are seen in Fig. 9 owe their
origin to the segmentation of the rim of the thin disc into which the
liquid has spread. The proceeding is probably exactly analogous to what
takes place in a sea wave that curls over in calm weather on a slightly
sloping shore. Any one may notice how, as it curls over, the wave
presents a long smooth edge, from which at a given instant a multitude
of jets suddenly shoot out, and at once the back of the wave, hitherto
smooth, is seen to be furrowed or "combed." There can be no doubt that
the cylindrical edge topples into alternate convexities and concavities;
at the former the flow is helped, at the latter hindered, and thus the
jets begin, and special lines of flow are determined. In precisely the
same way the previously smooth circular edge of Fig. 8 topples, and
determines the rays and lines of flow of Fig. 9.

Before going on to other splashes I will now endeavour to reproduce a
mercury splash of the kind I have described, in a manner that shall be
visible to all. For this purpose I have reduplicated the apparatus which
you have seen, and have it here so arranged that I can let the drop fall
on to the horizontal condenser plate of the lantern, through which the
light passes upwards, to be afterwards thrown upon this screen. The
illuminating flash will be made inside the lantern, where the arc light
would ordinarily be placed. I have now set a drop of mercury in
readiness and put the timing sphere in place, and now if you will look
intently at the middle of the screen I will darken the room and let off
the splash. (The experiment was repeated four or five times, and the
figures seen were like those of Series X.) Of course all that can be
shown in this way is the outline, or rather a horizontal section of the
splash; but you are able to recognize some of the configurations already
described, and will be the more willing to believe that a momentary view
is after all sufficient to give much information if one is on the alert
and has acquired skill by practice.

The general features of the splash that we have examined are not merely
characteristic of the liquid mercury, but belong to all splashes of a
liquid falling on to a surface which it does not wet, provided the
height of fall or size of the drop are not so great as to cause complete
disru