VOL. LXIX.j PHILOSOPHICAL TRANSACTIONS. 535 



it with some oppression, even from the beginning : which does not happen when 

 the inflammable air contained in a bladder is breathed, the lungs being in a natu- 

 ral state. 



And now it seems no longer difficult to give an answer to the question pro- 

 posed above, and to account for the small difference observed in the breathing 

 of the 2 different kinds of air in the bladders. The inflammable air, when 

 joined to a great quantity of common air, may be breathed safely, because there 

 is a quantity of common air sufficient for various inspirations, and that the 

 mixture of the 2 airs may be breathed till this common air is thoroughly in- 

 fected. But the inflammable air itself is not altered nor decomposed by the 

 respiration. We must therefore conclude, that the inflammable air is not such 

 a kind of air as can by itself alone be directly useful for respiration. It must 

 rather be considered as if there was nothing of that air in the case of the 

 bladder; and indeed it is found by experience, that the pulmonary air itself may 

 be breathed 8 or Q times in an empty bladder. The not being able to breathe 

 it ] 1 times successively, as was done when there was inflammable air in the 

 bladder, and the feeling an oppression in the first case and not in the second, 

 must be entirely attributed to the want of 35 cubic inches of air expirated, 

 which are necessary to give the lungs all the necessary expansion; whereas, in 

 the other case, the inflammable air serves to fill up space, and, together with 

 the common air, contributes to fill the lungs; so that the inflammable air, con- 

 sidered under these circumstances, and under this point of view, may be said to 

 be useful for animal respiration. This explanation seems most evidently demon- 

 strated by the following experiment. If 35 cubic inches of common air are 

 introduced into the bladder, and this air be breathed when the lungs are in a 

 natural state, it will be found that we may breathe it 20 times or more; whereas, 

 when the bladder was empty, it could not be breathed more than 9 times 

 at most. 



Before finishing this paper it will be proper to mention another cause, which 

 perhaps also contributes to render the inflammable air of the bladder less 

 noxious: this is the levity of the inflammable air itself with respect to common 

 air, which hinders the inflammable mixing with the common air. The 

 inflammable air swims continually on the common air, just as aether swims on 

 water; and the inflammable air swims still better than aether, because it is much 

 lighter in comparison than aether. Various experiments made on volatile sub- 

 stances have convinced me of this truth. If equal quantities of common and in- 

 flammable air, or dephlogisticated and inflammable air, be put into a tube, and 2 

 birds introduced in it, so that one of them may stand at the top, and the other 

 at the lower part of the inverted jar; it will be found, that the first of these 



