1877.] 



Forms assumed by Drops of Liquids. 



499 



stages was one by which the support was withdrawn from under a drop, 

 so as to leave it free to fall, while at the same instant an electrical circuit 

 was broken. 



The apparatus is drawn in the accompanying fig. 1. 



A B is a wooden rod about 6 inches long, turning about a horizontal 

 axis C C. When the rod is horizontal, a platinum wire D D, bound to 

 the underside, rests with one end on a platinum wire E E E, forming a 

 bridge, while the other end dips into a mercury-trough E, so that, with 

 the rod in this position, a current can pass from E to E, and so on 

 through the coils of the electromagnet of the relay described in my pre- 

 vious paper. 



The end B of the rod bears a little glass cup Gr made of the central 

 part of a watch-glass, covered with an adherent coating of smoke, ob- 

 tained by dipping the glass in paraffin-oil before smoking, on to which a 

 drop of water or mercury can be lowered from the vertical tube H, so as 

 to lie in the cup without adhering to it. A smart fillip with the thumb 

 knocks up the end A, and sends down the end B with the cup, thus 

 removing the support of the drop, and, of course, at the same time break- 

 ing the contact between D and E. K K is a w^ooden support bearing a 

 slip of card L, which acts as a spring-catch, and prevents the rod from 

 rebounding. The relay, described in the last paper, is then adjusted, so 

 that the primary spark of the E,uhmkorff's coil is obtained, and the drop 

 seen by it at any stage desired, whether in the air above the plate, at the 

 moment of impact, or later. 



A few precautions are necessary. 



The cup Gr must be smoked even with mercury, as that liquid adheres 

 slightly to glass, but not perceptibly to a smoked surface. With water 

 or milk it requires frequent resmoking. 



With a very large drop a cup of much deeper concavity must be used, 

 otherwise the drop lies flattened out on the cup ; and when the support 

 is removed, efficiency is given to the curvature at the sides, by the ten- 

 sion of which the lower part of the drop is driven downwards into a 

 column, which splits into two or more drops. 



I was also able to see the early stages with mercury by fixing the two 

 terminal wires of the electromagnet-current close under the end of a 

 vertical tube, from which the drops were allowed to fall directly on to 

 the plate. The convex end of the liquid, just before the drop fell, joined 

 the two terminals and allowed the current to pass, and when the drop 

 fell contact was broken. 



But the time of the spark was not quite so constant as in the method 

 just described, nor was the plan applicable to water, the conductivity of 

 which, even when strongly acidulated, was too slight. 



I noticed that, even when the spring of the relay was at its maximum 

 tension, and the point of the relay- wire only just in contact with the 

 mercury, the earliest spark that could be obtained showed that the drop 



