26 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 54 



0.11182 0.8S798 

 0.11204 0.88785 

 0.11202 



0.11204 0.88787, ± 0.000018 



0.11179 



0.11178 



0.11202 



0.11188 



0.11186 



0.11185 



0.11190 



0.11187 



0.11190, ± 0.000015 

 Dividing the mean of the second column by the mean of the first, we 

 have for the equivalent of oxygen : 



0.88787, ± 0.000018_ ^ 



0.11190, ± 0.000015- ^•^^^^' - "-""^^ 



Hence = 15.8690, ±0.0022. 



The details of the investigation are somewhat complicated, and involve 

 various corrections which need not be considered here. The result as 

 finally stated includes all corrections and is evidently good. 



The syntheses of water reported by Keiser ^ in 1898, involved the 

 direct oxidation of hydrogen occluded in palladium, with subsequent 

 weighing of the water so produced in the vessel in which it was gen- 

 erated. That vessel was tubular in form, and divided into two com- 

 partments; one containing phosphorus pentoxide, to absorb the water, 

 the other holding the palladium hydride. Each determination required 

 five weighings, as follows : First, of the vessel, containing only the diying 

 agent, and exhausted of air. Second, the same as the first, plus the pal- 

 ladium. Third, the gain in weight was measured after saturating the 

 palladium with hydrogen. Fourth, the entire apparatus after complete 

 oxidation of the hydrogen to water. The gain in weight gave the oxygen 

 absorbed. Fifth, like the fourth, but with the palladium removed. The 

 difference between the first and fifth weighings gave the amount of water 

 formed. All the operations were thus performed in a single piece of 

 apparatus, and troublesome corrections were avoided. The data obtained 

 were as follows, with weights not reduced to a vacuum standard : 



1 Amer. Chcm. Journ., 20, 733. 1898. 



