142 
PROFESSOR A. M. WORTHINGTON AND MR. R. S. COLE 
a very steady progress through the phenomenon., and in four test experiments that 
were made by photographing a solid sphere falling past a divided scale, three 
consecutive trials, made under ordinary good conditions, agreed within 3 - 0 V 0 second, 
while the fourth was not in error by as much as y-^o- second. 
When, however, the splash, not of a solid sphere, but of a falling drop was being 
observed, a further cause of irregularity was introduced.by the oscillations set up in 
the drop itself on its release, and by the slight adhesion between it and the 
supporting watch-glass. This adhesion is proved in the case of a wmter drop by its 
invariably carrying down with it a little lamp-black from the smoked surface. A drop 
of milk, on the other hand, carried down very little, and on this account, and probably 
also because of the greater viscosity of milk, the splash of a drop of milk is less 
troublesome to follow in its initial and most rapidly changing stages than is that of a 
drop of water. We had also reasons to suspect that after setting a drop in place on 
the watch-glass the film of intervening air gradually escaped and led to a suctional 
adhesion if the release were too long postponed. Dusting the watch-glass, after 
smoking, with lycopodium powder appeared to diminish the adhesion. 
Nevertheless, when care was taken to preserve regularity in the procedure, the 
same one of us always manipulating the Wimshurst machine and the laboratory 
releasing key, and the other the dark-room releasing key and the setting of the drop 
in place, the apparatus worked and worked well, and if the steps taken between 
the photographs of a series are as much as to^o of ^ second, reversals of the proper 
order will be exceptional, and there is no difficulty in obtaining stages at closer 
inteiwals if desired. 
We found, however, that between series of photographs taken on different days 
there were sometimes noticeable breaches of continuity in the timing, which may be 
attributed to changes in the potential difiereiice between P and Q, and therefore in 
the distance from them of the timing sphere wdien the flash took place, and perhaps 
to other causes that escaped our notice; we have consequently distinguished photo¬ 
graphs taken on different days by letters placed just above the right-hand corner. 
For the rest it may be observed that it has not yet seemed worth while to press 
the accuracy of the timing much beyond what is required for a complete record of 
the consecutive phenomena. 
No misapprehension can arise as to the times assigned, if it is remembered that 
they refer to the setting of the apparatus, and are liable to such uncertainties as have 
been mentioned. 
The Photographs. —Series I. (Plate 1), consisting of 33 photographs, gives the splash 
of a water drop, weighing '2 of a gram.,* falling 40 centims. into milk mixed with water, 
* Drops of a constant size were obtained from a vertical glass tube, connected by indiarubber tubing, 
with a wide funnel, in Avhicb the level was slightly higher than the month of the delivery tube. The 
experimental drop was made up of a definite number (1, 2, 3, or 4) of such drops, caught in a smoked 
and lycopodium-dusted watch-glass and thus conveyed to the I’eleasing cup. 
