GASEOUS AND LIQUID STATES OF MATTER, 
577 
vacuum was formed till about one-fourth of the gas was removed. On restoring the 
pressure, a column of mercury entered and occupied the place of the expelled gas. By 
withdrawing the end of the tube from below the surface of the mercury in the capsule, 
and again exhausting cautiously, the column of mercury could be reduced to any 
required length. The tube, when thus filled, had the form shown (figure, p. 576). 
Two file-marks had been made, one at d , the other at e, in the narrow part of the 
tube, about 10 millims. distant from each other, and the capacity of the tube from a 
mark near a to d, and also from the same mark to e, had been determined by filling it with 
mercury at a known temperature and weighing the mercury. The tube was now placed 
accurately in a horizontal position and connected by an air-tight junction with one limb 
of a long U-tube filled with mercury. Each limb of the U-tube was 600 millims. long, 
and 11 millims. in diameter. By removing mercury from the outer limb of the U-tube, 
a partial vacuum was obtained, and the column of mercury (mn) was drawn into the 
narrow tube ( df ). From the difference of capacity of this part of the tube, the column 
of mercury was now about four times longer than before. It was easy with a little 
care so to adjust the pressure that the inner end of the mercurial column coincided with 
the mark e. When this was accomplished, the difference of level of the mercury in the 
two limbs of the U-tube was accurately read by means of a cathetometer, and the height 
of the barometer as well as the temperature were carefully noted. Similar observations 
were made with the gas expanded to the mark d. Two independent sets of data were 
thus obtained for calculating the volume of the gas at 0° C. and 760 millims., and the 
results usually agreed to less than x&oo P ai 'h The tube, after being disconnected with 
the U-tube, was cut across a little beyond e, as shown (figure, p. 576), and was now 
ready to be introduced into the pressure apparatus. 
The capillary tubes were calibrated with great care, and their mean capacity was 
determined by weighing a column of mercury whose length and position in the tube 
were accurately observed. Gne millim. of the air-tube used in these experiments had 
an average capacity of 0-00002477 cub. centim., and 1 millim. of the carbonic-acid-tube 
of 0-00003376 cub. centim. A table was constructed showing the corrected capacity of 
each capillary tube from the sealed end for every millimetre of its length. An allowance 
of 0-5 millim. was made for the cone formed in sealing the tube. 
For the sake of clearness I have described these operations as if they were performed 
in the detached tube. In actual practice, the tube was in the brass end-piece before it 
was filled with gas (Plate LXIII. fig. 7). 
The construction of the apparatus employed in these experiments will be readily 
understood from figures 3 and 4, Plate LXIII., which exhibit a section of the simple form. 
Two massive brass flanges are firmly attached round the ends of a cold-drawn copper 
tube of great strength, and by means of these flanges two brass end-pieces can be 
securely bolted to the ends of the copper tube, and the connexions made air-tight by 
the insertion of leather washers. The lower end-piece (fig. 7) carries a steel screw, 
180 millims. long, 4 millims. in diameter, and with an interval of 0'5 millim. between 
