the Densities of Gases and Vapours, 339 



interposed between the glass and the iron frames. This junc- 

 tion, which is made once for all, soon becomes hermetically 

 tight. 



The experiment is conducted in the following manner. 



The sheet-lead casing with its transparent glass is raised from 

 the sheet-iron base p q s after the screws x, x, x have been taken 

 out. The tubes AB and CD are adjusted in their tubulures. 

 Fig. 14, which represents a magnified vertical section of the piece 

 cdef, shows sufficiently the mode of adjustment. I will merely 

 say that the tube B c or the tube C D is coated with a plait of 

 hemp well stuffed with red-lead cement, and the plait is strongly 

 screwed on the annular spaces o o by means of the screws K, L. 

 The tubes AB, CD ought to be quite parallel. They are ad- 

 justed in a vertical position by moving the levelling-screws V, V. 



I assume that the density of a gas is to be determined at 

 various temperatures and under different pressures. 



The stopcock R being in the position 3 (fig. 8), mercury is 

 poured into the open tube C D, and a vacuum made by means of 

 the air-pump which is fixed to the stopcock r. The vacuum is 

 made several times, dry air being admitted each time. The 

 object of this is to make the insides of the tube A B completely 

 dry. The stopcock B is then gently turned into the position 1 

 (fig. 8), while the vacuum is still produced in the tube AB; 

 the mercury from C D passes into the tube A B. The stop- 

 cock r is closed when the level of the mercury reaches the sum- 

 mit of the capillary tube A b. The tube bm is connected with 

 a bell-jar full of the gas whose density is to be determined, or 

 with the apparatus which disengages it in a state of purity; 

 the stopcock r is opened, and mercury allowed to flow out from 

 the tube AB by placing the stopcock R in the position 4 (fig. 8). 

 The s'topcock r is closed, and the stopcock R placed in the posi- 

 tion 1 (fig. 8) as soon as enough gas has been introduced into 

 the tube A B ; the quantity varies according to the conditions of 

 compression or of expansion under which the gas is to be studied. 



The tubes A B and C D are divided into millimetres. In a 

 preliminary experiment, the capacities of the tube AB corre- 

 sponding'to the divisions traced on the tube have been determined 

 with great accuracy, by weighing the quantity of mercury which 

 fills these capacities, and successively allowing them to run out 

 by the stopcock R placed in the position 2 (fig. 8). The casing 

 EFGH is now adjusted on the sheet iron p qs, and this case 

 is filled with water at the temperature at which the experiment is 

 to be made. This water is kept in continual motion by a stirrer, 

 and its temperature is maintained quite uniform. By allowing 

 mercury to flow out by the stopcock 11, or by adding it to the tube 

 C D, the same quantity of gas may be successively put under the 



2 A2 



