1S25.] Atmospheric Air by Hydr<ygen, 305 



cent, of oxygen. All the results below must be considered a» 

 averages of four or five experiments. 



- ^^^ y^ahulan^^Arrcingement of the Experiments. 



N. B. Those experiments marked h denote that hydrogen remained in the residue 

 after detonation ; those marked o denote oxygen ; and that marked n denotes that 

 neither of the two was found. 



On this table it may be remati:edn' I. The V^Tf-transition 

 from no detonation to a perfect one, when the oxygen is near 

 a minimum; 2. The slow transition at the bottom of the table 

 when the hydrogen is near the minimum; and 3. That there is 

 no imperfect combustion about the middle of the table ; either 

 the oxygen or the hydrogen, or both, are always entirely gone. 

 But there is one anomaly that calls for explanation, — the neu- 

 tralizing proportions of oxygen and hydrogen appear to be 

 t^'to^ 2-1, when it is well known that they are nearly, if no6 

 exactly, as 1 to 2. This is occasioned, no doubt, by the impu- 

 rity produced in the hydrogen ; first, in the reception of it in 

 bottles with water charged with atmospheric air ; and, secondly' 

 in the subsequent passing of it two or three times througlr 

 water in the process for detonation. That this is the true reasotf 

 is cijnfirmed by theloss being only 3, and not 3'1 in that (ias'e^;^ 



iS/eic Series f vol. x. x 



