AND RATIO OF THEIR ATOMIC WEIUIITS. 



113 



Dittmar and Henderson,* Rayleigli,f and by Leduc.;}'. The values which they 

 found are given in the following table, together with the value computed above 

 from the ratio of the densities ; they are arranged in the order of their magnitude. 



SUMMARY OF DETERMINATIONS OF THE ATOMIC WEIGHT OF OXYGEN. 



The result of Berzelius and Dulong has so large a probable error that it may 

 be disregarded. Those of Dumas and of Erdmann and Marchand may be safely 

 dismissed from consideration, on account of the error detected by Dumas at the 

 end of his work. There remain only the results which have been published within 

 the last ten years. Six of these results differ but little from their mean, which has 

 the value 15.88: one result differs from this mean by 0.07 ; it is difficult to resist 

 the suspicion that some source of constant error vitiated this discordant value. The 

 results which remain are those upon which our present knowledge of the atomic 

 weight of oxygen will doubtless be considered to rest. They ai-e repeated, with 

 some details which may serve to assist in judging of the trustworthiness of the sev- 

 eral series of determinations, for the convenience of any who may not care to refer 

 to the original papers. 



VALUES ON WHICH OUR KNOWLEDGE OF THE ATOMIC WEIGHT OF OXYGEN DEPENDS. 



Observer. 



Dittmar and Henderson. 



See First Table. 

 Cooke and Richards. 

 Morley. I. 



II. 



III. 

 Leduc. 

 Rayleigh. 

 Noyes. 



§ ■" ti 



&^a 



Oxygen and water 



Ratios 



Hydrogen and water 



Hydrogen and water 



Ratios 



Hydrog;n and oxygen. 



Oxygen and water 



Hydrogen and oxygen 



Hydrogen and oxygen 



s s 



u 



7 (of II) 

 32 and 47 

 16 

 II 

 23 and 10 



12 



2 



5 



24 



4) T^ 

 g O j= 



See First 

 0.42 gr. 



3-75 ■' 

 To 3.75 gr. 



3-75 gr- 



o.ii " 

 0.77 " 



21.4 gr. 

 Table. 



To 31 gi 

 30 gr. 

 18.7 •■ 

 0.9 " 

 6.1 " 



* Proceedings of the Philosophical Society (Glasgow), 22, t,^ (1891). 

 t Proceedings of the Royal Society, 45, 425- 



X Comp/es Petidus, 115, 43. § The probable error of Leduc's result is 

 assumption that one of his experiments was equal in accuracy to one of 



o :: £" 



6 « •; 



24. 1 gr. 



3-8 " 

 34 " 



15.S66 

 15.868 

 iS.Sftg 

 15.879 

 15-879 

 15.879 

 15.881 



15.89 

 15.897 



0.0062 



0.0017 



0.0017 



0.00066 



0.0014 



0.00032 



0.012 



0.0088 



0.0017 



arbitrarily assigned on the 

 Dittmar and Henderson's. 



