of Vapours to Mariotte and Gay-Lussac's Law. 287 



was indicated by fine Geissler normal thermometers graduated to 

 the tenth part of a degree, which were controlled by comparison 

 with other normal thermometers, and by repeatedly checking their 

 fixed points. By means of a double stirrer, which could be rapidly 

 moved np and down, a uniform temperature was preserved 

 throughout the bath. Outside the bath a T-shaped glass tube, 

 pom n, was now connected with the protruding end of the tube, 

 c d, by an india-rubber tube and some luting- wax. The descend- 

 ing branch (o n) of this tube, which was provided with a perforated 

 glass cock, was connected with an air-pump, while the other end, 

 om, conducted into a chloride-of-calcium tube, u, and thence 

 into "a manometer, ef. The connexion between these last two 

 ends was made by means of an india-rubber plug, which em- 

 braced the narrow tube and was forced into the larger one. All 

 the points of connexion were so tightly secured that the appara- 

 tus, so long as it was in use, was perfectly air-tight, even under 

 the highest ranges of the manometer. A barometer (g h) of a 

 very wide bore gave the atmospheric pressure, whilst a thermo- 

 meter (t') placed beside this and the manometer gave the corre- 

 sponding temperature. 



The course of investigation was as follows : — As soon as the 

 calibrated tube, being perfectly dry, was filled with warm, very 

 pure and dry mercury, freed to the utmost from air, and when a 

 bursting bulb containing a weighed quantity of fluid had been 

 placed on this, it was closed by means of a small wooden 

 disk, lined on one side with caoutchouc and provided with a 

 knob on the other, and being then inverted was placed in the 

 larger iron sheath. The latter operation was rendered possible 

 by placing about the sheath a wooden case which, filled with 

 mercury, afforded plenty of room for the purposes of manipula- 

 tion. Into the smaller iron sheath the tube cd was introduced 

 half filled with mercury. The remaining half of the same, being 

 still free, served for the reception of the mercury that over- 

 flowed from the calibrated tube in the course of the experiment, 

 while the circumstance that the lower half already contained 

 mercury facilitated the necessary compression. Eor compres- 

 sions, the calibrated tube a b, as well as the india-rubber collar 

 embracing it, was secured firmly to the iron sheath by means of 

 iron rods and a cross tie. This portion of the apparatus being 

 thus prepared was placed in the bath, and, with the principal 

 tube in a strictly vertical position, was united, after the fashion 

 described above, to the other part, which was fastened to a strong- 

 fixed table on which the whole stood. Now, to measure the 

 mass of air from which such a large tube could scarcely be kept 

 entirely free, the air in the intermediate part of the apparatus 

 (dp omue) was greatly rarefied by means of the air-pump while 

 the bath was kept at a given temperature ; and after closing the 



