104 JOURNAL OF THE 



evidence of the struggle to retain the old unit hydrogen, and at 

 the same time secure accuracy for it, bv fixing; definitely its relation 

 to oxygen, by means of which nearly all comparisons with the 

 other elements must be made. I fear that much of this work 

 would not have been done if it had been made to appear that the 

 atomic weight of hydrogen, and not of oxygen, was aimed at. 

 The atomic weight of hydrogen has only the ordinary interest 

 of that of any of the elements. With the atomic weight of 

 oxygen they all stand or fall, no matter which we choose as our 

 theoretical unit. The new determinations referred to are: 



Rayleigh (Chemical News, Vol. 57, p. 73), 0= 15.9 12. 



Cooke & Richards (Am. Chem. Journal, Vol. 10, p. 81), 

 0=15.953. 



Cooke & Richards (Am. Chem. Journal, Vol. 10, p. 91), 

 0=15.869. 



Keiser (Am. Chem. Journal, Vol. 10, p. 250), 0= 15.949. 



Morley,* is at work upon it and cites Scott, H:0 :: 1.994: 1, 

 by volume. This makes the atomic weight of oxygen less that 

 16, as all of the others do. None of them come up, indeed, to the 

 present assumption of 0=15.96, so if these investigators are 

 correct, or more nearly correct than those who have preceded 

 them, the whole table of atomic weights must be again shifted. 



It is time, then, that hydrogen were finally discarded as the 

 unit. Oxygen is in every way preferable, and nothing like uni- 

 formity will be attained until it is adopted. 



With oxygen as the standard, what value shall be assigned it? 

 Four values have been suggested: 1st, 0=1 ; 2d, O=10; 3d, 

 0=16; 4th, O=100. 



If 0=1 we would have nearly 10 per cent, of the elements 

 represented by decimal fractious and the following partial table 

 would show other inconveniences: 



Al. 1.694 Cr 3.2^4 



F 1.194 Fe 3.501 



Mg 1.500 Co 3.67 



Na 1.4408 Cu 3.95 



P 1.940 Mn 3.4:'. 



S 2.0037 Ni 3.67 



Ti 3.15 



Si 1.754 V 3.20 



♦American Chemical Journal, Vol. 10, p. 21. 



