226 MAGNUS ON THE EXPANSIVE FORCE OF STEAM. 
mercury then fell suddenly, the watery vapours were disengaged, 
and, as their expansive force was now greater than that of the 
rarefied air, they threw the mercury violently out of the tube 
abd. Not only was the experiment rendered wholly valueless, 
but the tube a 4 d had to be replaced by a new one; and some- 
times the mercury was projected with such force through the 
whole apparatus that it penetrated into the air-pump, so that 
this had to be taken immediately to pieces to preserve it from 
injury by amalgamation. To protect the air-pump from similar 
mishaps, I had a bulb blown in the tube gk at A, in which the 
tube k h terminated in a point curved upwardly. In this bulb 
the whole of the mercury must collect, however violently it might 
be projected into the tubes. I could however find no means of 
preventing the adhesion of the mercury and water in the boiled 
tube abd. I introduced into it before it was curved a piece of 
polished iron wire, this floated after the mercury had been boiled 
and projected into the water, but even with this wire the same 
phenomenon resulted, only when a little bubble of air had col- 
lected above the water have I not observed it. I shall however 
subsequently recur to some conclusions drawn from this phe- 
nomenon. 
If the mercury had sunk in the closed arm, some time elapsed 
before it acquired a constant position, evidently because the in- 
ner space had been covled by the latent heat requisite for the 
formation of steam, and only gradually re-acquired the tempera- 
ture of the surrounding medium. If the expansive force of the 
vapour was constant, the height of the air thermometer was 
read off by means of the kathetometer, then by means of the 
same instrument the difference between the heights in the tube 
abd in the case, as well as the height of the water in the closed 
arm, and likewise the heights of the mercury in the gauge o p q, 
and the barometer 7 s ¢, fig. 1 and 4, measured. 
One great advantage of the apparatus described consists in its 
being not only applicable to the measurement of pressures lower 
than the pressure of the atmosphere, but likewise for such as 
are higher. In this case it is only requisite to arrange the air- 
pump so that we can not only exhaust with it, but likewise con- 
dense, which the air-pumps commonly in use in this country, 
which are constructed with the so-called Grassmann’s cock, 
readily admit of. The air is then condensed previously to the 
warming of the case S P R, and when the temperature has be- 
