266 Mr. F. Claudet on a Class of 



making a difference only of ()-37 per cent. The results obtained, 

 liowever, agree pretty well together ; and as they do not differ 

 much from the calculated numbers, I am much inclined to believe 

 sixteen the true number of equivalents of hydrogen in the salt ; 

 and I am further confirmed in this view by the manner in which 

 the salt is decomposed by heat. A combustion-tube about two 

 feet long was closed at one end and bent at right angles within 

 about half an inch of the closed end, so as to form a kind of re- 

 tort. A certain quantity of the salt was rubbed into a paste 

 with a little water and rolled up into the size of a pea. When 

 quite dry, this was dropt into the tube and made to enter the 

 small retort; mercury was then gently poured into the tube, 

 which was gradually filled and then inverted in a mercurial 

 trough. The mercury descended about a quarter of an inch in 

 the tube, on account of a small quantity of air which remained 

 in that portion containing the salt. The retort part of the tube 

 was now slowly heated by means of a spirit-lamp until the salt 

 was entirely decomposed. The gas produced occupied nearly 

 the whole of the tube, which was two feet in height. On allow- 

 ing the tube to cool, and introducing a small quantity of hydro- 

 chloric acid, the whole of the gas was absorbed with the excep- 

 tion of a column of about three-quarters of an inch in height, 

 showing that the space above the mercury was entirely composed 

 of ammoniacal gas. Now the decomposition of this salt into no 

 other gas than ammonia, and no other solid products than chlo- 

 ride of ammonium and protochloride of cobalt, is only compatible 

 with a certain number of atoms of hydrogen, which is sixteen, 

 for — 



3Cl,2Co,5N,16H=:2CoCl + NH4Cl-F4NH3. 



Had there been one or two equivalents less of hydi'ogen, one 

 equivalent of ammonia would have been broken up, giving hy- 

 drogen and nitrogen not condensed by the hydrochloric acid. 



Assuming, then, the above number of atoms to be correct, and 

 applying Berzelius's theory of the copulated compounds, the for- 

 mula of this salt may be written — 



3(NH4Cl)+2(NH2Co); 



that is, a compound of 3 equivalents of chloride of ammonium 

 with 2 equivalents of an ammonia, in which 1 atom of hydrogen 

 is replaced by cobalt. In fact the salt has the characters of 

 such conjugate compounds. It has the properties of chloride of 

 ammonium with regard to form and taste ; while on the other 

 hand the basic property of the 2 equivalents of ammonia have 

 totally disappeared, the salt being quite neutral to test-paper. 

 This compound is analogous to the remarkable platinum com- 

 pounds discovered by Gros and Reiset ; but with this difference. 



