344 Mr. C. Akin on a New Method for the Direct Determination 



where 7 is the real specific heat of the gas, or its specific heat 

 under constant volume, and g its weight, we obtain 7, as referred 

 to the specific heat of water = 1, 



The magnitude l is measured directly by the thermometer, and 

 6 is known by means of formula (I.) Similarly, r may be measured 

 directly by a timepiece, and i and i y by means of a galvanometer. 

 In order to determine also r and r x accurately, the following 

 method might be advantageously adopted. Having ascertained 

 ?, let the spiral be omitted from the circuit of the battery, but 

 be replaced by a length of standard wire, of standard tempera- 

 ture, sufficient to reduce the intensity of the current to its ori- 

 ginal value i. If / be the length of standard wire required, then 



In like manner r l may be obtained, 



r i = h> 

 where ?, is known ; so that in the end 7 also becomes entirely 

 known. 



3. Having thus exemplified the principle of the method in an 

 assumed case, we may now proceed to state in what manner the 

 conditions implied in that example may be, very nearly at least, 

 practically realized, or otherwise superseded by equivalent arrange- 

 ments. The first condition which it is necessary to fulfil is, to 

 preserve the volume of the gas constant during heating. For 

 this purpose two slightly different means may be suggested. 

 First, let A in the adjoining figure re- 

 present a spherical vessel of rigid sub- 

 stance filled with gas, and B a siphon or 

 manometric tube filled with mercury,from 

 a up to b. If c be a very small hole in the 

 tube near its upper extremity, and of suffi- 

 cient tenuity, the column of mercury, in 

 ascending through the very small range 

 between b and d, will not be able to pass 

 through the hole c } although it may be 

 possible for air to pass freely through 

 the hole. Hence, though the gas in A 

 be heated, its volume will remain unaltered as soon as the mer- 

 cury column has reached the top d of the tube B, which may be 



* It is supposed in the above formula that the resistance of the remainder 

 of the circuit is not perceptibly influenced by the interchange of the two 

 resistances represented by r and /. This might practically be secured with 

 sufficient approximation. 



