680 APPENDIX. 



When an oil bath is used there is more difficulty in obtaining the 

 same temperature in the vapour and the bath. When the oil 

 bath is heated to within 20 or 30 of the point at which we wish 

 to stop we must damp the fire. This causes the temperature to 

 rise more slowly. When we are within 8 or 10 of the point 

 the fire must be drawn. This causes the increase of temperature 

 to be very slow, and enables that of the vapour to become as high 

 as that of the oil. 



The balloon is now removed from the bath, and wiped clean 

 with the greatest care. When cold and clean it is weighed. 

 The increase of weight gives the quantity of matter in the balloon 

 that had been converted into vapour. 



The beak is now plunged into mercury and the point broken 

 off. The mercury enters the balloon and fills it completely, if 

 the whole air had been expelled. If not, a portion of air remains, 

 the volume of which must be noted and subtracted from the ca- 

 pacity of the balloon. When the experiment has been properly 

 made the residue of air does not exceed Ol or 0-2 cubic inch. 



The capacity of the balloon is determined by filling it with 

 mercury, and measuring the mercury by pouring it into a gra- 

 duated vessel. By these determinations we know the weight of 

 the vapour and its bulk, from which we deduce the specific gra- 

 vity. The excess of the weight of the balloon full of vapour, 

 minus the weight of the air which the balloon contains, gives us 

 the weight of the vapour. 



Knowing the volume of the balloon and the temperature of 

 the air when it was weighed, we bring this volume by calculation 

 to what it would be, supposing the thermometer at 32 and the 

 barometer at 30 inches, and this corrected volume is converted 

 into weight, by the known weight of 100 cubic inches of air at 

 32, and when the barometer stands at 30 inches; namely, 

 32-79 grains. 



As the balloon was increased in bulk by the high temperature, 

 we must calculate how much that was, and allow for it. This is 

 easily done, as we know that the expansion of glass for 1 of Fah- 

 renheit is T s JS T . All these corrections being made we have the 

 weight and the bulk of the vapour ; and, dividing the latter by 

 the former, the quotient gives us the specific gravity of the va- 

 pour under examination. 



Now let us see how this knowledge of the specific gravity of 



