in a Recalculation of the Atomic Weights. 107 



Sb = 119*955 ; + *036. In view of the recent discussions upon 

 the subject, we may regard the lower figure as established. 



Bismuth. — General mean from Schneider's experiments 

 Bi = 207*523; +-082. If = 16, this becomes 208-001. 

 All of Dumas's results on bismuth chloride give a mean of 

 Bi = 210*464; the figures which he. considers best give 

 Bi = 209*78. Schneider's work is probably nearest correct, 

 his method being less liable to constant errors than that of 

 Dumas. 



Tin. — Berzelius, Mulder and Vlaanderen, Dumas, and 

 Vlaanderen. General mean, Sn = 117*698 ; +*040. 



Titanium. — Rose's weighings give Ti = 48*710 + *105. 

 Mosander's figures give Ti = 47*045. Pierre's giveTi =49*889 

 + '096; and Demoly's, Ti = 52*191 + 153. A general mean 

 of the results of Pierre, Rose, and Demoly is 49*846 + *064. 

 Mosander's work must be rejected for want of sufficient de- 

 tails. An experimental revision is wanted. 



Zirconium. — From Berzelius's figures, Zr = 89*255 ; + '039. 

 From Marignac's data Zr = 90*328; +*113. The general 

 mean of both sets is Zr = 89*367 ; +'037. New determina- 

 tions are evidently needed. 



Thorium. — Berzelius, Chydenius, Berlin, Delafontaine, 

 Hermann, and Cleve. General mean, Th= 233*414 ; +'073. 



Gallium.— From Boisbaudran's results, in mean, Ga = 

 68*854. 



Indium. — Reich and Richter, Winkler, and Bunsen. Gene- 

 ral mean, In= 113*398 ; ±*047. 



Cerium. — Beringer, Hermann, Marignac, Bunsen and 

 Jegel, Rammelsberg, Wolf, Wing, and Buehrig. General 

 mean of all, Ce= 140*424; +*017. Buehrig's analyses of 

 the oxalate give Ce = 141*198 ; + *020. The figures by Wolf 

 and by Wing give Ce = 137*724. Wolf and Wing had a 

 white ceroso-ceric oxide ; and Wolf suspects the ordinary yel- 

 lowish compound to contain a fourth metal of the cerium group. 

 Buehrig's work is the best ; but the possible presence of a 

 fourth metal is not considered by this chemist. Therefore 

 new experiments are needed. 



Lanthanum. — Hermann, Rammelsberg, Marignac, Czud- 

 nowicz, Holzmann, Zschiesche, Erk, and Cleve. General 

 mean, La= 138*526; + -030. 



Didymium. — Marignac, Hermann, Zschiesche, Erk, and 

 Cleve. General mean, Di = 144*573 ; +*031. Cleve's work 

 alone, which is doubtless the best, gives Di = 146*804. 



Scandium. — Nilson's results, recalculated, * give Sc = 

 43*980; ±*015. 



Yttrium. — Popp, Delafontaine, Bahr and Bunsen,and Clove. 



