CHEMISTRY FOR 1887 AND 1888. 



By F. W. Clahkk. 



GENERAL AND TIIEORETIOAL. 



Atomic iceifjhts ; oxygen. — The composition of water and the relation 

 between the atomic weights of hydrogen and oxygen have been discussed 

 •luring the year by Cooke and Richards, Scott, Keiser, Morley, and 

 Lord Rayleigh. 



Cooke and Richards adopted the following plan : Hydrogen, puritied 

 with extraordinary care, was weighed in large glass globes which had 

 previously been exhausted by means of a Sprengel pump. It was then 

 burned by passing it over heated coi)per oxide, and the water so formed 

 was weighed. Three series of experiments, numbering sixteen in all, 

 were made, and distinguished from each other by the use of hydrogen 

 derived from different sources. First, hydrogen was used which had 

 been evolved from zinc and sulphuric acid; secondly, hydrogen was 

 obtained by electrolysis ; and in the third series the gas was derived 

 from the action of potassium hydroxide upon aluminium foil. In this 

 way the errors due to possible impurity in hydrogen from only one source 

 were sought to be avoided. The mean results by series were as follows, 

 wlum H = l : 



Series I, five experiments 0=15. 0")! ; ^ .0048 



Series II, five fxperinients If). 951^) ; -J- .0022 



Series III, six experiments lo.9.")2 ; J^ .0035 



Average of the sixteen experiments 15.953; -f^ .0017 



The quantity of hydrogen burned in each experiment was a little over 

 0.4gramnu'. The oidy i>()s.sil)le impurity in it was nitrogen, which was 

 looked for spectroscoi»ically, but not I'ound. Its absence however could 

 not be certaiidy proved, and its effect, if present, would be to render the 

 apparent atomic weight of oxygen too low; in other words, its ])resence 

 would lead to possibly erroneous conclusions relative to " Trout's law." 

 (Proc. Amer. Acad., xxiii, 140.) 



The investigations of Alexander Scott relate to the compo.sition of 

 water by volume. It is well known that oxygen and hydrogen, as a(;t- 

 ually measured at ordinary temperatures and ])ressures, do not rigidly 

 conform to F'>oyle's law. They vary frotn it in opposite directions ; and 

 this fact must be taken into account before we can deduce their relative 



425 



