188 
ME. W. EAIEBAIEX A2sT) 3IE. T. TATE OX THE DEXSITT 
globe, form the saturation-gauge to which reference has been made. So long as the 
steam in the globe A remains in a state of saturation, the inner column remains stationary, 
at a point a little above the level of the outer column, so as to balance the column of 
water in the steam-bath. But when, in raising the temperature, the whole of the water 
in A is evaporated, the steam becomes superheated, and then the pressure in A no 
longer balances the pressure in the steam-bath B, and the inner column of mercury 
rises and the outer one falls, the difference of level forming a measure of the expansion 
of the steam. It was found a matter of the greatest importance that the obseiwer should 
not in these experiments trust to the unaided eye to determine the point at which the 
columns began to diverge, but that a careful series of measurements of the difference of 
level should be made, not only near the saturation point, but also at various tempera- 
tures of superheating ; thus affording data for determining the law of expansion near 
the saturation point, and for estimating the maximum temperature of saturation from a 
point at which any error arising from the cohesion between the water and the glass, or 
from the retention of portions of water in the steam itself, might be eliminated. For 
reasons hereafter assigned, it was also found advisable to take these readings of the 
levels of the columns rather in a descending than in an ascending series of temperatiu-es. 
To read the column levels with rapidity and facility, seeing that they could not be 
approached within six or eight inches, a simple form of cathetometer was derised, suffi- 
ciently accurate for the purpose, and much more convenient than the complicated 
French instrument sometimes employed for this purpose. This cathetometer is shown in 
elevation, in Plate VIII. tig. 2 ; a is a heavy iron base resting on three levelling screws ; 
into this is screwed firmly the planed iron stem c c, and adjusted by the levelling screws 
in a vertical position ; on the stem slides a brass block, h, carrying a small telescope, 1 1, 
with sight wires, and fixed by a thumb-screw behind. Fig. 3 shows a part of the stem, 
c c, with the graduation into inches and tenths, the block h and vernier v moving with 
the telescope, by which the levels of the columns were read off to hundredths of an 
inch. In Plate VII. fig. I, ff shows the position of the cathetometer in the experiments, 
on its iron base or pedestal, g. 
The steps in the process for determining the specific grarity of steam by this appa- 
ratus were as follotvs : — 
A glass globule, of a size to contain as nearly as might be the required quantity of 
water for vaporization, was selected from a series similar in form to those sliomi in 
fig. 6, Plate VIII. In order to avoid the collision produced by the bursting of a sealed 
globule of water, as in Gay-Lussac’s method, these globules had open stems, and after 
being filled with water were immersed hot in a cup of mercury (fig. 7), so that in cooling 
the mercury should rise and fill in part the capillary stem. The weight {il\) of the cup. 
mercury and globule, all perfectly dry, was first ascertained by means of a delicate 
balance ; the globule being then filled with water, the stem plunged in the mercury, and 
the surface of the mercury being dried, the weight (w.^) of the whole was determined. 
In this state the cup of mercury was transferred, and tlic globule passed into the large 
