Error in Calorimetric Work. 325 



between the two instruments being known throughout the 

 extent of their scales. 



It was found that one hour * was a sufficient interval to 

 allow for the salt to assume the temperature of the surrounding 

 atmosphere, provided this latter remained constant during 

 this time. In order to diminish the error in measuring the 

 rate of cooling (or heating) during the dissolution of the salt, 

 the temperature of the air was always maintained as nearly 

 as possible at the temperature of the calorimetric water, the 

 difference in the majority of cases not exceeding '1° C, or at 

 the most '2° C. The correction to be applied, owing to the 

 temperature of the salt not being exactly the same as that 

 of the calorimetric water, was thus rendered very small, and in 

 most cases quite negligible. 



The alteration in the temperature of the salt during 

 its removal to the calorimeter can scarcely be appreciable, 

 since this operation does not take more than ten or fifteen 

 seconds. 



The calorimeter employed was an open calorimeter similar to 

 that devised by Berthelot (Mdcanigue Chimique, vol. i. p. 140); 

 a pounder such as Berthelot uses to crush salts which cake 

 (ibid. vol. i. p. 183) being employed in all cases, not only as a 

 pounder, but also as a stirrer, and the platinum vessel in which 

 the operations were performed being always used without its 

 cover ; indeed, free circulation of air round the calorimeter is 

 most important in all delicate determinations. It was found 

 that an annular cardboard disk, covering in a most imperfect 

 manner the space between the calorimeter and the surround- 

 ing water-jacket, enclosed the air to an extent sufficient to 

 make it act as so much calorimetric water, and introduced a 

 very appreciable error into the results. 



In dissolving an anhydrous salt, when caking generally 

 occurs, the calorimeter proper cannot be placed on the cork 

 tripod as in other cases, but must be supported on a cork disk 

 covered with cardboard ; and, inasmuch as this disk was found 

 to absorb an appreciable amount of heat from the calorimeter, 

 the quantity thus absorbed had to be determined by perform- 

 ing a series of experiments in which a reaction of known 

 calorimetric value was effected in the calorimeter when sup- 

 ported on the tripod and on the disk alternately. The u:e of 

 this disk unfortunately introduces a small experimental error, 

 since from continual pounding the form of the bottom of the 



* A hygroscopic substance may be left in a tube well closed with an 

 india-rubber stopper for two or three hours with safety ; but a longer 

 exposure should be avoided, as moisture is gradually absorbed from the 

 air permeating the stopper. 



