540 



SCIENCE PROGRESS. 



90*8296 of the bromide. These give us the following 

 values : — 



Stas 



van der Plaats 



Thomsen 



Ag. 



NH4CI. 



107-930 53"53i9 

 107-9244 53-5291 

 107-9299 53-5318 



CI. 



35-457 



35H5 6 5 



35*4494 



NH4. 



18-0749 

 18-0726 

 18-0824 



I 



N. 



HifO = i6 



I 



i4 - 055 



14-0519 



14-0396 



1-005 



1-0052 



1-0107 



OifH = 1. 



or 15-920 

 or 15-917 

 or 15-830 



Similarly we have — 



Stas 



van der Plaats 



Thomsen 



NH 4 Br. 



98-0324 

 98-0273 

 98-0323 



Br. 



79-95 2 



79-9548 

 79-9510 



NH 4 . 



18-0804 

 18-0725 

 18-0813 



H ifO = 16 Oif H = i. 



1-00635 

 1-00515 

 1-01042 



15-899 

 15-918 



I5-835 



The final results are therefore not very concordant, 

 ice using Stas' own numbers, O = 15*899 to 15*920, 



& 



with van der Plaats' (21) recalculated values of Stas' atomic 

 weights we get 15*917 to 15*918, with Thomsen's (22) re- 

 calculated values of the same, 15*830 to 15*835. 



One conclusion at any rate seems certain, and that is, 

 that if H = 1, then O = 16 is too high, thus corroborating the 

 work previously detailed above. In 1883 Professor Dewar 

 and the author (23) attempted to use the substituted am- 

 monias so as to throw additional light on this subject, but 

 the difficulty of obtaining absolutely pure triethylamine, free 

 from traces of methyl and the higher radicals, prevented 

 the results from being as conclusive as one would desire. 

 The value 182*012 was obtained for (C. 2 H 5 ) 3 NHBr; 

 subtracting Stas' value for ammonium bromide from this, 

 we get 182*012 — 98*032 = 83*98 for C 6 H 12 , and taking 

 C = 12003 we g" et H 12 = 11*96 or H = "996. Recently a 

 similar number has been obtained for hydrogen by Pro- 

 fessor Thomsen (24) of Copenhagen, by means of ammonia 

 and hydrochloric acid, but by a totally different method. 

 He passes dry hydrochloric acid gas into a weighed flask 

 containing - water coloured with litmus and again weighs 

 the flask, this gives the amount of hydrochloric acid 

 absorbed. Dry ammonia gas is now passed in until the 



