OF ARTS AND SCIENCES. 175 



endeavoring to perfect this larger apparatus, and a very thorough 

 knowledge was acquired of its relative efficiency. The greatest gain 

 that could have been expected in carrying out the work on this scale 

 would have been the reduction of the probable error to about one 

 half of the present amount, but it is obvious that this gain could be 

 of no importance in the present condition of the science. The accu- 

 racy we have reached is far beyond the demands of any analytical 

 work ; and, as we have shown, the theoretical question in regard to 

 Front's law has been settled so far as analytical work can solve the 

 problem. It now turns solely on the typical character of the material 

 we call hydrogen, when prepared in the purest condition known to 

 modern science. 



In considering the bearing of the result now published on Prout's 

 hypothesis, it must be borne in mind that it confirms in a most strik- 

 ing manner the result of Dumas, based on the weight of oxj^gen which 

 water contains, and in connection with his results furnishes a com 

 plete analysis of water, with a degree of accuracy as great as can be 

 expected, or as has ever been obtained, in any analytical work. 



Complete Analysis of Water. 



Percentage of Oxygen after Dumas 88.864 ±0.0044 



Percentage of Hydrogen according to the present work 11. 140 ±0.0011 



100.004±0.0045 



It must be remembered that in Dumas's investigation the oxygen 

 alone was weighed, while in the present investigation the hydrogen 

 alone was weighed, and the fact that these two wholly independent 

 analvtical results made under such widely different circumstances 

 exactly supplement each other within the limits of probable error, is 

 an evidence of accuracy and a proof of finality which is irresistible. 



It would have been highly desirable, if it had been possible, to deter- 

 mine both the oxygen and the hydrogen in one and the same analytical 

 process, as the writer succeeded in doing in the case of silver, bromine, 

 and antimony, and he made many experiments on the reduction of 

 oxide of silver by hydrogen with this view. He succeeded in pre- 

 paring pure oxide of silver, of definite composition, but the investi- 

 gation was interrupted by the failure of his sight before he was able 

 to overcome the grave experimental difficulties which he process 

 presented. In view, however, of the present results, it is doubtful 

 whether any advantage would have been gained by that mode of 



