OF ARTS AND SCIENCES : MARCH 8, 1864. 257 



and ill-adapted bulbs, had so much of the oxide of copper reduced that 

 combustible gasses passed through the absorbing apparatus ; and in one 

 instance, when the unabsorbed gases were collected, the quantity of 

 combustible gas was so considerable as to form, with the oxygen col- 

 lected with it, — which of course came forward at an earlier or later 

 stage of the process, — an explosive mixture. Having seen no indica- 

 tions that any other than gaseous bodies escaped the combustion tube 

 in such a case, it occurred to me that such an analysis might be saved 

 by collecting the gas over mercury, and, at the close of the combustion, 

 before detaching the absorbing apparatus, conducting it a second time 

 through the combustion tube.* As a matter of economy, also, in the 

 saving of the excess of oxygen, when a considerable number of analyses 

 are to be made, this idea seemed to recommend itself; as the oxygen 

 would, at the same time, become purified from any traces of combustible 

 matter which might be present, and could then safely be collected as 

 pure oxygen, and finally transferred to the oxygen gasometer. 



I therefore constructed for this purpose the apparatus which is repre- 

 sented in the background of the accompanying figure as attached to 

 the anterior end of the absorption apparatus. At the close of the com- 

 bustion, when only pure oxygen appears to enter the potash bulbs, the 

 flow of oxygen is interrupted ; the communication with that portion of 

 the drying apparatus which is back of the short U tube, A, is closed at 

 b ; and the tube B — which is movable in the cork — turned up.f The 

 joint at c is then disconnected; the end leading to the receiver C 

 tightly closed with a piece of glass rod ; and a communication estab-: 

 lished between the absorption apparatus and another receiver containing 

 water, — not shown in the figure, — for collecting the pure oxygen. 

 On opening the spring-clip d (the more modern form, which is provided 

 with a fine screw, is excellently well adapted for this purpose), the 

 mercury will flow from the reservoir D into the receiver C, and force 



* As the time consumed in an experiment is so short, and the quantity of 

 combustible gas present, if any, so very small, and that mixed with a very large 

 quantity of oxygen, it is not improbable that the gas might as well be collected 

 over water; as the quantity which could be absorbed by the water in so short a 

 space of time would probably be inappreciable. 



t That this tube may not operate as a siphon, the outer limb is formed by 

 attaching, near the bend, a flexible tube, of larger bore than that of the glass tube. 

 Tliis flexible tube is preferable to glass, on account of the readiness with which it 

 adapts itself to any change of position of the glass tube, by which it may always 

 project into the receiver underneath, and prevent waste of mercury. 



