106 Prof. F. W. Clarke on the Results obtained 



Iron. — Berzelius, Svanberg, and Norlin, Erdinann and 

 Marchand, Maumene, Rivot, and Dumas. General mean, 

 Fe = 55*913; +-012. 



Copper. — Berzelius, Erdmann and Marchand, Millon and 

 Commaille, and Hampe. Cu = 63*173 ; +'011. 



Molybdenum. — Debray, Dumas, Rammelsberg, and analyses 

 by Liechti and Kemp. Mo = 95*527 ; + *051. 



Tungsten. — Berzelius, Schneider, Mareliand, v. Borch, 

 Dumas, Bernoulli, Persoz, Roscoe, Scheibler, and Zettnow. 

 General mean of all, W = 183*610 ; + "032. 



Uranium. — Peligot, Ebelmen, and Wertheim. General 

 mean, U = 238*481; + *082. 



Aluminum. — Berzelius, Tissier, Isnard, Dumas, Terreil, 

 and Mallet. General mean, Al = 27*0092 ; ±'0028. All 

 the data except Mallet's might be rejected without essentially 

 affecting this value. 



Gold. — Berzelius and Levol. Au=196*155 ; + *095. A 

 thorough redetermination is much needed. 



Nickel. — Several early series of nickel and cobalt were re- 

 jected. The data taken were by Schneider, Dumas, Russell, 

 Sommaruga, Winkler, and Lee. General mean, M = 58*547 ; 

 + •009. I attach more importance to the concordant results 

 of Schneider, Sommaruga, and Lee, whose figures give a 

 general mean of Ni = 57*928 ; +*022. Lee's investigation 

 was of all the least susceptible to constant errors. 



Cobalt. — Schneider, Dumas, Russell, Sommaruga, Winkler, 

 Weselsky, and Lee. Co = 58*887 ; +*008. 



Selenium. — Berzelius, Sacc, Erdmann, and Marchand, Du- 

 mas, and Ekman and Pettersson. Se = 78*797; +*011. 



Tellurium. — Berzelius, v, Hauer, and Wills. General mean, 

 Te = 127-960; +*034. The results of v. Hauer and Wills 

 upon K 2 TeBr 6 give Te = 127*170; ±*173. Wills's mini- 

 mum figures give me Te = 126*07. In all of these results 

 certain constant errors are possible ; so that the question 

 raised by Mendelejeff as to whether tellurium is above or below 

 iodine, cannot be regarded as settled. 



Vanadium. — Roscoe's weighings, recalculated with the new 

 values for 0, Ag, and CI, give Y=51*256 ; +*024. 



Arsenic. — Pelouze, Dumas, Wallace, and Kessler. General 

 mean, As=74*918 ; +*016. Wallace's analyses of AsBr, 

 were made to establish the atomic weight of bromine ; but 

 they serve a better purpose here. 



Antimony.— For this element there have been the two rival 

 figures 120 and 122. The general mean from data by Kessler, 

 Dumas, and Dexter is Sb = 122*092; +*035. The general 

 mean from the experiments of Schneider and of Cooke is 



