230 MAGNUS ON THE EXPANSIVE FORCE OF STEAM. 
It may appear surprising that the observations present such 
great variations ; but if it be considered that 01 C., at least 
with temperatures above 50° C., causes a difference in the ex- 
pansive force of about 2™™, it would scarcely be possible to ex- 
pect greater accuracy, The readings of the air thermometer 
may indeed be obtained to within 0°01 C., but the errors pos- 
sible in these readings render such slight differences of tempera- 
ture uncertain. Besides such errors dependent on the observa- 
tion of the temperature, the method itself gives rise to the 
following. In the first place, the coincidence of the horizontal 
wire of a telescope with the meniscus of mercury is always un- 
eertain on account of the reflexion, and then in the carrying of 
the telescope along the kathetometer its axis neyer remains per- 
fectly parallel. This latter inconvenience I have rendered less 
great by always adjusting this telescope too high, and clamping 
it to the kathetometer, and then gradually lowering it by means 
of the micrometer screw until the line touched the mercurial 
trough ; in fact, by always operating with the micrometer screw 
in the same direction in adjusting the telescope. With this 
precaution the sum of the errors of all the seven readings be- 
longing to one observation did not amount at the utmost to 
0°15™™, within which limit all the experiments made at 0° come, 
Moreover, the observations might be rendered inaccurate from 
my want of success in depriving the water employed entirely of 
air, either by interrupted and long-continued boiling, or even by 
allowing it to ascend almost boiling hot through the mercury. 
Already Watt, and after him Southern, experienced the same 
difficulty *. No air was, it is true, visible at the ordinary tem- 
perature and pressure of the atmosphere, but when vapour had 
formed by the diminution of the pressure and increase of the 
temperature, an air bubble remained above the water after it had 
cooled, which however, under the pressure of the atmosphere, was 
only about 0°5™™ in diameter. After some hours it had always 
disappeared and was absorbed by the water. When the experi- 
ment was again commenced the above-mentioned sudden evolu- 
tion of steam usually happened. 
Since the space which was occupied by the vapour in the 
tube abd was never smaller in my experiments than 7 cub. 
cent., the air bubble occupied at 0°, and the pressure of one 
atmosphere at the furthermost 0:00001 of this space, the expan- 
* Robison’s System of Mechanical Philosophy, ii. p. 31 and 170. 
[ Sarees 
