MAGNUS ON THE EXPANSIVE FORCE OF STEAM. 223 
ments, intimating that the above-mentioned apparatus might 
easily lead to errors; I therefore felt it necessary to test it again 
before employing it in the present researches. 
With this view I had made a second air thermometer of ex- 
actly the same construction as the former, with the exception 
that it contained the air in a vessel curved in the shape of a fork. 
The vessels of both thermometers were inserted in the apparatus 
in question, so that the one was exactly situated between the 
other forked one, and with it in the same horizontal plane, as 
exhibited in fig. 3. The following numbers are the tempera- 
tures which the two thermometers indicated on the contempo- 
raneous application of heat. 
Air Thermometer. 
Vessel with two arms, | Vesse: with one arm. Difference. 
° ° ° 
158-138 158-18 +0-05 
187-79 187-96 +0:17 : 
220-73 220-67 —0-06 
287-60 288-14 +054 | 
From this it is evident that the temperatures coincide as much 
as could be expected in these higher portions of the scale, con- 
sequently I did not object in the least to employ this apparatus 
for the contemporaneous heating of the vapours and air thermo- 
meter. It presents only one very great evil, which is, that it 
always requires several hours before the temperature in the inner 
case becomes constant, so that the experiments require a very 
great sacrifice of time. 
The arrangement in which the vapour was generated con- 
sists of a short tube bent in the form of a U, and about four 
inches long, which is represented in a b de, fig. 2. One of its 
extremities is closed, and in order to increase the space for the 
vapour blown out at a to a bulb. At the other extremity a 
glass tube 0 c is affixed at 6 at right angles, which passes through 
the envelope P of the case required for the heating and project- 
ing from this as far asc. The short closed arm of this tube is 
filled with mercury which is well-boiled, then some water which 
had previously boiled violently without intermittence from half 
to three quarters of an hour was poured through c into the open 
arm, and a small quantity of it while still warm conveyed by in- 
Sh It is searcely necessary to mention, that in this case, on the application 
of heat, a tension mtist occur in the glass, and consequently an itregular ex- 
pansion of the bulb. 
R 2 
