on the Atomic Weight of Carbon. 223 



The weighings as well as the burnings were conducted in 

 a thin crucible of Meissner porcelain, which was covered by 

 a platinum lid. All our experiments were fortunately per- 

 formed with the same crucible, the weight of which therefore 

 was controlled whenever weighings were made. 



The calcination of the acetate of silver proceeds very 

 easily, without either swelling up or spirting. The salt at 

 first becomes gray, and, on cautiously heating acetic acid, 

 distils off, and then the salt assumes a brown colour; at 

 length, when the odour of the acid is no longer observable, 

 there remains a gray skeleton of silver, which retains the form 

 of the salt burnt. If the heat be now increased, and the lid 

 raised, so as to admit a current of air, a visible glowing is 

 observed throughout the whole mass, and there remains a 

 spongy mass of shining white metallic silver. 



After cooling, the crucible with the silver is weighed, heated 

 to redness anew, weighed again, and so on, until not the 

 slightest change of weight is exhibited. The absence of car- 

 buret of silver was always particularly demonstrated by dis- 

 solving the silver in dilute nitric acid. 



The weighings were effected with a balance, which indi- 

 cated quite distinctly half a millegramme, even when loaded 

 with 20 grammes. The weights (made by Oerthing of Berlin) 

 were carefully compared previously to employment, and were 

 found to exhibit no appreciable variation in their subdivisions. 

 Lastly, we adopted the precaution of conducting one-half of 

 the experiments with newly prepared salt, by which means 

 the fact, that the numbers yielded by salts prepared by dif- 

 ferent operations agree, is rendered very obvious. 



The tartrate of silver is not easily obtained in a crystalline 

 form. When the nitrate of silver is precipitated in the cold 

 by a solution of pure Rochelle salt, from which, by the addi- 

 tion of a little nitric acid, the weak alkaline reaction which it 

 generally possesses has been removed, a caseous and not a 

 crystalline precipitate is formed. On the other hand, if the 

 precipitation is effected by mixing boiling hot and dilute solu- 

 tions, the liquid becomes brownish, but not turbid, and on 

 cooling, metallic silver in brown laminae falls down. Again, 

 when -a dilute solution of nitrate of silver is heated to from 

 80° to 85° C, and to this a hot concentrated solution of the 

 tartrate of potass and soda is added, a precipitate falls, which, 

 by agitating, at first disappears ; if the addition of the tartrate 

 of potass and soda be discontinued as soon as the precipitate 

 is permanent, and is not redissolved, fine scales of the tartrate 

 of silver separate on cooling, which, after being well washed 

 and dried, are very white, and have a metallic lustre, resem- 



