206 SCIENTIFIC APPARATUS. 



means of the information thus acquired, to fix the exact composi- 

 tion of a body under investigation. All that is necessary is to 

 determine the weight of a known volume of the gas or vapour, 

 but the process of course varies with the condition of the body at 

 the ordinary temperature. It a gas, three processes may be em- 

 ployed. The most accurate one is that employed by Regnault, in 

 which he used a flask, of ten litres capacity, which could be con- 

 nected by means of brass stopcocks to an air-pump, a pressure 

 gauge and a receiver containing the gas of which the specific 

 gravity was to be determined ; the tube between the receiver and 

 the globe being provided with apparatus for thoroughly drying 

 the gas. After the flask was surrounded with ice, it was exhausted 

 by the air-pump and filled with dry gas from the receiver. This pro- 

 cess was repeated several times, so as to ensure the purity of the 

 gas in the globe. When the globe was filled and the pressure of 

 the enclosed gas measured, the stopcock was closed, and the flask 

 suspended from the pan of a large but delicate balance, to the 

 other pan of which a globe of equal size was attached. After 

 weighing, the globe was once more surrounded with ice, as much 

 gas as possible pumped out, the pressure of the remainder mea- 

 sured, and the flask again weighed. The difference between the 

 two weighings gave a number, from which the weight of the gas 

 contained by the flask could be calculated. The same series of 

 operations was then carried out with the standard gas (hydrogen 

 or air), and from the results obtained the specific gravity was 

 determined. 



This process is extremely delicate and adapted for standard 

 determinations ; a simpler one, which can be performed more 

 rapidly, and giving results sufficiently accurate for all ordinary 

 purposes, is that devised by Bunsen. 



A third process for the determination of the specific gravity of 

 gases was devised by Bunsen, and is based on the discovery by 

 Graham that the velocities with which gases pass through a minute 

 hole in a thin metallic plate their rates of effusion are inversely 



