1 1 •_> DENSITIES OF OXYGEN AND HYDROGEN 



of the experiments are given, with some facts which may assist in funning an 

 opinion of the merits of the experiments, in the folhiwing table : 



RATIO OK DENSITIES OF OXYGEN ANU HYDROGEN. 



The hist fohinm gives weights assigned according to the probable error,s, which 

 are given in the previous column. The mean, with these weights, is 



1) = 15.S87 ± 0.0017. 



Since the weights assigned to the observations at common temperatures are five 

 times as much as those assigned tlic observations at 0" C, we may consider this 

 mean as holding for common temperatures ; we therefore combine with it the value, 

 given by Scott* from forty-seven selected observations, of the ratio of the volumes 

 in wliich the gjises combine when measured at common temperatures, 



R =: 2.CO245 ± 0.0005. 



From these latios we t)btain the atomic ratio, 



O = 15.868 ± 0.0017. 



The atomic ratio has been determined directly by Berzelius and l)ulong,t 

 Duma.s,J Krdmaiiii and Mai-chand,§ Cooke and Richards,! Keiser,^ Noyes,** 



* Philosophical Traiistif lions of the Royal Society, 184, A, 543 (1893). 



f Annales tie Chiinie ct Je Physique, [2] 15, 389. 



\ Annales Je Cliimie el de Physique, [3] 8, 200. 



§ Journal fiicr /raktisclie Chemic, 26, 468. 



11 American Chemical 'yoiirnal, 10, Si, and 191. 



•i American Chemical your nal, 10, 249. 



*• American Chemical journal, 12, 441. 



