1513.1 On the Specific Gravity of the Gases, 1 79 



rained without adding any thing whatever to the precision of the 

 results. Thus the chance of error is increased, without any 

 advantage. 2. It is more difficult to determine the specific 

 gravity of hydrogen gas with accuracy than that of any other 

 gas. It is so light that the presence of the vapour of water 

 makes a sensible alteration in the result. Accordingly that 

 weight varies according to the care with which the gas has been 

 dried. Hence to employ hydrogen as a standard would be not 

 only to subject the result to all the chances of inaccuracy from 

 the experiment itself, but to all those likewise to which hydrogen 

 itself is exposed. Tiie specific gravity of common air can be 

 taken with more ease, and with m.ore accuracy, than that of any 

 other gas whatever : nor have I any doubt that the knowledge of 

 its specific gravity resulting from the experiments of Sir George 

 Shuckburgh is as precise as is wanted for the most delicate 

 investigations. 



The weight of 100 cubic inches of each gasj as given in the 

 table, is supposed to have been taken when the thermometer 

 stood at 60°, and the barometer at 30 inches. 



I shall now proceed to state the documents according to which 

 the preceding table was drawn up. 



1. The weight of 100 cubic inches of dry atmospherical air 

 was determined by Sir George Shuckburgh Evelyn. His method 

 was to ascertain the relative weights of water and air, and then 

 to determine the exact weight of a cubic inch of water. Now 

 a cubic inch of distilled water at the temperature of 60" (when 

 the corrections made upon Sir G. Shuckburgh Evelyn's experi- 

 ments by Mr. Fletcher [See Nicholson's Journal, vol. iv. 

 p. 35] are attended to) weighs 252*506 grains : and the weight 

 of air is to that of water at the temperature of 60°, and when 

 the barometer stands at 29i inches as 0*001188 to 1, according 

 to Sir G. Shuckburgh. Hence if the barometer were at 30 

 inches (the thermometer remaining at 60°) the weight of air 

 would be to that of water as 0*001208 to 1. From this it is 

 easy to deduce, by the common rule of proportion, that if 100 

 cubic inches of water weigh 25250-6 grains, 100 cubic inches of 

 air will weigh very nearly 30*5 grains. If we take the French 

 estimate of the weight of a cubic inch of water at 60°, namely, 

 252*72 grains, the weight of 100 cubic inches of air will turn 

 out nearly 30*53 grains ; but as the French experiment was 

 made at the temperature of 40°, and the change upon the bulk 

 of water when heated up to 60° is made by calculation, there is 

 a greater chance of error in such an estimate than in calculating 

 from the experiments of vSir G. Shuchburgh Evelyn, where the 

 heat differed only a few degrees frpm 60°. On that account I 

 consider his result as the one which chemists ought to adopts 



