PRESSURE ON THE TEMPERATURE OP VOLATILIZATION OP SOLIDS. 
41 
be 4‘16 millims., of which 3’59 millims. are clue to the tension of vapour in the 
condenser, while the difference between 4T6 and 3’59:=0 - 57 millim. must be due to 
air. Since this pressure of air is nearly constant throughout the experiment, the 
slight variation being due to difference in temperature of the apparatus, which can 
be approximately allowed for, it is possible to calculate from Regnault’s table the 
variation in temperature of the ice caused by altering the temperature of the condenser. 
For instance, the lowest temperature of the condenser in series I. is — 11°*1 ; this 
corresponds to a pressure of 1'905 millims. ; the total pressure in the apparatus is 
therefore 1‘905 + 0‘570=2 , 475 millims., and the corresponding temperature is — 7 0, 9 ; 
the temperature observed being —8 o, 0. 
In the second series the pressure of the air found as above is 1'08 millims., the 
lowest temperature being taken as the basis of calculation. The agreement between 
the calculated and found differences is seen to be close until the bulb became exposed, 
when its temperature rose more rapidly. 
8. As it was sufficiently proved by these experiments, which were confirmed by 
numerous others, that our hypothesis is fairly in agreement with experimental 
evidence, it was decided next to admit small quantities of air, and to ascertain what 
rise in temperature the ice underwent. But as a very small amount of air causes 
a great difference in temperature, we were unsuccessful in measuring the exact 
amount of each addition. In order to do this, it was necessary to have as nearly as 
possible a perfect vacuum in the apparatus. For this purpose a litre of distilled 
water was boiled down to about 200 cub. centims. and introduced, while almost at its 
boiling-point, into the apparatus. The thermometers were not inserted until steam 
was issuing rapidly from the three orifices. Before inserting them they were held for 
some time in the steam, so as to remove any adhering film of air. It was evident that 
the bulbs were nearly vacuous, for two columns of water enclosing vapour came com¬ 
pletely together without showing any trace of a bubble of air. It is right here to 
observe that all attempts to produce a complete vacuum with the Sprengel’s pump 
which we then possessed were totally unsuccessful 
After the apparatus was in order the freezing mixture was changed, and the 
temperature of the ice fell to —17°; the temperature of the thermometer in the 
condenser, however, fell very slowly, owing possibly to absence of convection 
currents. The temperature of the ice rose, while that of the condenser fell, until at 
— 13° they were equal, and remained constant for a considerable time. It thus 
appears that the temperature of volatilization of the ice is —13° when the pressure in 
the apparatus is equal to the vapour-tension of ice at that temperature. A minute 
quantity of air was then introduced ; the temperature of the condenser was now 
—12 0- 6, while that of the ice rose at once to — 6 0, 7, and then remained stationary. On 
a second addition of air, with the condenser at —12 0, 9, the temperature of the ice rose 
to —1 0, 5 Thus, with no air, the difference between the readings of the two ther¬ 
mometers was 0° ; after the first addition, 5 0- 9, and after the second addition of air, 11° - 4. 
MDCCCLXXXIV. 
G 
