80 ATOMIC WEIGHT DETERMINATIONS. 



difference, 2.15. Ag = 107.66; S = 31.98; 01 = 35.37; 

 O = 15.96. The specific gravities of the chlorides not hav- 

 ing been determined, the weighings are not reduced to 

 vacuum. The pentachloride was prepared from M by heat- 

 ing it in a current of 01 entirely free from air. The metal 

 had been freed from oxide by heating in an atmosphere of 

 HCl. By moderate heating of the pentachloride in dry H, 

 and by distilling pentachloride over the product in dry 

 carbon di-oxide, the trichloride is obtained. The trichlo- 

 ride heated in carbon di-oxide is decomposed into tetrachlo- 

 ride and di-chloride, which latter must be purified with 

 warm dilute nitric acid. {Liebig's Ann., 169, 1874, 360, 

 344.) 



NIOKEL. 



Regnault has determined the specific heat of nickel. It 

 corresponds to an atomic weight of about 59. {Gmelin- 

 Krauty I. c.) 



E. RoTHOFF : 69.09 (O = 16) ; 369.333 (O = 100). 



Rothoff converted 188 parts of oxide into chloride, a 

 neutral solution of which gave 718.2 parts argentic chlo- 

 ride. If 01=221.64, Ag= 1349.66, the value follows. 

 {Berzelius' Lekrbuch, S, 1221.) 



. P. Berthier. 



Lassaigne having announced the atomic weight of nickel 

 at 500, {Schweigger's Jahrbuch, 9, 108,) Berthier re-ex- 

 amined the subject and found Eothoff's number confirmed. 

 {Berzelius^ Jahresbericht, 5, 1825, 148 ; Annul, de Chim. et de 

 Fhys., 25, 1824, 148.) 



Erdmann and Marchand : 58.2 ( = 16); 365.9 (O 

 = 100). 



Determined "with all precaution " by the reduction of 

 the oxide with hydrogen. The results varied from 29.1 to 

 29.3, but Erdmann has reason to believe the smaller number 

 the more accurate. [Erdmann's Journ. fiir Prak. Chem., 

 55, 1852, 202.) 



