ON BOILING WATER. 
107 
num wire, about 1J inch long, and bent horizontally into a curve. Distilled water, 
which had been well boiled and cooled under the receiver of an air-pump, was poured 
into this flask so as to fill about one-fourth of its capacity. It was then placed under 
the receiver of an air-pump, and one of the copper wires brought in contact with a me¬ 
tallic plate covering the receiver, the other bent backwards over the neck of the flask, 
and its end made to rest on the pump-plate. By this means, when the terminal wires 
from a voltaic battery were made to touch, the one the upper and the other the lower 
plate, the platinum wire would be heated, and the boiling continued indefinitely in the 
vacuum of a very excellent air-pump. The effect was very curious; the water did not 
boil in the ordinary manner, but at intervals a burst of vapour took place, dashing the 
water against the sides of the flask, some escaping into the receiver. (There was a pro¬ 
jection at the central orifice of the pump-plate to prevent this overflow getting into the 
exhausting tube.) 
After each sudden burst of vapour, the water became perfectly tranquil, without a 
symptom of ebullition until the next burst took place. These sudden bursts occurred 
at measured intervals, so nearly equal in time, that, had it not been for the escape 
from the flask, at each burst, of a certain portion of water, the apparatus might have 
served as a timepiece. 
This experiment, though instructive, did not definitely answer the question I had pro¬ 
posed, as I could not of course ascertain whether there was some minute residuum of 
gas which would form the nucleus for each ebullition ; and I proceeded with others. A 
tube of glass, 5 feet long and T %ths inch internal diameter, was bent into a V shape ; 
into one end a loop of platinum wire was hermetically sealed with great care, and the 
portion of it in the interior of the tube was platinized. When the tube had been well 
washed, distilled water, which had been purged of air as before, was poured into it to 
the depth of 8 inches, and the rest of the tube filled with olive oil; when the V was 
inverted, the open end of the tube was placed in a vessel of olive oil, so that there would 
be 8 inches of water resting on the platinum wire, separated from the external air by 
a column of 4 feet 4 inches of oil. The projecting extremities of the platinum wire 
were now connected with the terminals of a voltaic battery and the water heated; some 
air was freed and ascended to the level of the tube—this was made to escape by care¬ 
fully inverting the tube so as not to let the oil mix with the water—and the experiment 
continued. After a certain time the boiling assumed a uniform character, not by such 
sudden bursts as in the Florence-flask experiment, but with larger and more distinct 
bursts of ebullition than in its first boiling. 
The object of platinizing the wire was to present more points for the ebullition, and 
to prevent soubresauts as much as possible. 
The experiment was continued for many hours, and in some repetitions of it for days. 
After the boiling had assumed a uniform character, the progress of the vapour was 
carefully watched, and as each burst of vapour condensed in the oil, which was kept 
cool, it left a minute bead of gas, which ascended through the oil to the bend of the 
tube: a bubble was formed here which did not seem at all absorbed by the oil. This 
was analysed by a eudiometer, which I will presently describe, and proved to be nitrogen. 
The beads of gas, when viewed through a lens and micrometer scale at the same height 
in the tube, appeared as nearly as may be of the same size. No bubble of vapour was 
condensed completely, or without leaving this residual bubble. The experiment was 
frequently repeated, and continued until the water was so nearly boiled away, that the 
oil, when disturbed by the boiling, nearly touched the platinum wire ; here it was neces¬ 
sarily stopped. 
To avoid any question about the boiling being by electrical means, similar experiments 
were made with a tube, without a platinum wire, closed at its extremity, and the boiling 
was produced by a spirit-lamp. The effects were the same, but the experiment .was 
more difficult and imperfect, as the bursts of vapour were more sudden, and the duration 
•of the intervals more irregular. 
The beads of gas were extremely minute, just visible to the naked eye, but were made 
visible to the audience by means of the electric lamp. 
In these experiments there was no pure boiling of water, i. e. no rupture of cohesion 
of the molecules of water itself, but the water was boiled, to use M. Donny's expression, 
by evaporation against a surface of gas. 
It is hardly conceivable that air could penetrate through such a column of oil, the 
