

Numerics 



of the Elements. 







Group XIV. 







^ = 210 



-15(-9375) z . 







X. 



y- 



y calc. 



Au . 



. . 1 



196-16 



195-94 



Os . 



? 







Hg • 



. . 6 



199-71 



199-82 



Tl . 



. . 13 



203-72 



203-52 



Pb . 



. . 22 



206-47 



206-37 



Bi . 



. . 28 



207-52 



207-54 





Group XV. 







j/ = 225- 



-15(-9375) x . 







X. 



y- 



y calc. 



Ng . 



. . . 14 



218-93 



218-92 





Group XVI. 







r = 240 



-15(-937S)*. 







X. 



y- 



y calc. 



Tho. 



. . . 10 



232-37 



232-13 



U . 



. . . a 



239-70 



240-00 



397 



On the Specific Values of some Numerics. 



Nitrogen. — The details of Stas' recent work have not yet 

 reached me. His latest result is 14*053. 



Magnesium. — If we reject Dumas' determination (Clarke, 

 p. 107), the mean of the older results is 24*19. The mean of 

 this and Marignac's (1884), Mg = 24*37, has been accepted. 



Potassium. — Stas' new result is stated as 39*142. 



Manganese. — The best of the older work, in my opinion, is 

 that of Berzelius and Turner : mean 54*86. Marignac (1884) 

 gives 55*07. The mean of these numbers has been accepted. 

 Dewar and Scott (1883) obtained 55'04. 



Chromium. — Baubigny (1884) finds a still nearer value, 

 viz. 52*16, 



Nickel and Cobalt. — In spite of the very great labour 

 already expended on these numerics, they still remain open 

 to question. I have given Clarke's averages, which have, as 

 will be observed, a place in the theory. 



Copper. — Ever since the publication of Wright and Menke's 

 " Notes on Manganese Dioxide " [Chem. Soc. J. (trans.) 

 1880, p. 22], I have felt grave doubts as to the exactness of 

 the composition of higher oxides. The numeric of copper 

 has been decided by several series of analyses of the higher 

 oxide, confirmed by electrolysis of the sulphate. Both of 



