330 Dr. Thomas Thomson on the Atomic Weight, Spc. 



an atom of protoxide of nickel weighs 4*25, and an atom of 

 peroxide 4*75, we have, 



Protoxide of nickel . . . 24-67 or 4-218 



Sulphuric acid 29-24 or 5* 



Water ....... 46-09 or 7-881 



100-00 



The water of crystallization, in this case, is very near 

 seven atoms, exceeding that number by only 0*006 or only 

 y-Lpth part of an atom. I think the reason of this small 

 anomaly, making the atom of protoxide of nickel 4-218, 

 instead of 4-25, is probably, that a small portion of the 

 oxide of nickel examined, may have existed in the state of 

 protoxide. For I find that by long exposure to an intense 

 red heat, the oxygen in the black oxide of nickel is sensibly 

 diminished. I once succeeded in getting an oxide by 

 intense heat of an ash gray colour, with a very sliglx^ tint 

 of green, which was reduced by means of hydrogen gas in 

 the way above described. 5-05 grains of it sustained a loss 

 of 1-19 grains. Hence, it was composed of. 



Nickel 3-86 or 3-25 



Oxygen 1-19 or 1-001 



5-05 



so that it was very nearly in the state of protoxide. But, a 

 strong and long continued heat is requisite to bring it to 

 this state. 



From the experiments stated in this paper, which were 

 made with every possible attention to accuracy, no doubt 

 remains on my mind that the atomic weight of nickel is, 

 as I long ago stated, 3*25, and thatRothoff's number 3-716 

 is the consequence of his having taken the peroxide of 

 nickel for the protoxide. I must, therefore, retract the 

 statement made in the Records, (vol iii. p. 255,) and admit 

 that the error lies not in the atomic weight of nickel, but 

 in the number given by Dulong and Petit for the specific 

 heat of that metal. If the specific heat of nickel be 0-1 15, 

 instead of 0*1035, then the law of Dulong and Petit would 

 apply with accuracy to that metal, supposing its atomic 

 weight to be 3*25, as I believe it to be. 



