Chemical Affinity in terms of Electromotive Force. 333 



iodine per 500 cub. centims. of gas produced, representing a 

 diminution in the volume of gas evolved practically quite in- 

 appreciable. The gas-measuring apparatus in our possession 

 being incapable of measuring so large a quantity as 500 cub. 

 centims., we employed the method used by Joule for deter- 

 mining the volume of gas produced, viz. collection over water 

 in a bottle, the weight of which was known when filled with 

 water and closed by a stopper. To determine the gas collected, 

 the bottle was immersed for some minutes in a large bucket 

 of water, the temperature of which was known ; when the gas 

 had attained the temperature of the water, the bottle was 

 raised until the level of the fluid inside and out was the 

 same (*". e. until the pressure was atmospheric) ; the stopper 

 was then inserted and the bottle removed, wiped dry, and 

 weighed. In this way a close approximation to the quantity 

 of gas produced was obtained, the error being one of defect, 

 owing to the sources above mentioned and the absorption 

 of oxygen by the water during the process of collection. From 

 the number of cub. centims. of mixed gases thus obtained 

 (reduced to dryness, 0° C, and 760 millims.) =v, the weight of 

 water decomposed, w, was calculated by the formula 



w=vx 0-0005363, 

 the coefficient 0*0005363 being deduced from Eegnault's ob- 

 servations that 1 litre of dry hydrogen and oxygen at 0° and 

 760 millims. weigh respectively 0*089578 and 1-429802 

 gramme, 



whence 2 grammes of hydrogen occupy 22 "327 litres, 

 and 15*96 „ oxygen „ 11*162 „ 



so that 17*96 grammes of detonating gas occupy 33*489 „ 



1 7*Qf* 



or 1 cub. centim. of detonating gas weighs ooXfto" 0*0005363 

 gramme. 664z * J 



40. The calorimeter employed in these experiments con- 

 sisted of a glass beaker capable of holding about 1500 cub. 

 centims., fitted closely inside a polished tin cylinder supported 

 (on a wooden block and several folds of wadding) concentri- 

 cally inside a similar tin cylinder, so that about an inch of air- 

 space everywhere intervened between the two tin surfaces, the 

 two tin cylinders being kept in the same relative positions by 

 cork wedges. This outer cylinder was weighted with lead and 

 sunk inside a third similar larger cylinder filled with water; so 

 that the calorimeter itself was surrounded on all sides, saving the 

 top, by a water jacket. Polished tin lids for the calorimeter and 

 the outermost vessel were provided, perforated with holes for 

 the passage of the voltameter (in the axis of the cylinders), 



