PUTEKFACTIOlir AND INFECTION. 47 



that air which has passed through the lungs is known 

 to have lost its power of causing putrefaction. Such air 

 may mix freely with the blood of an internal wound 

 without risk of mischief; and that truly great scientific 

 surgeon had the penetration to ascribe this immunity 

 from danger to the filtering power of the lungs. Prior 

 to my becoming acquainted with this hypothesis in 

 1869, I had demonstrated its accuracy in the following 

 nner.' 

 Condensing in a dark room, and in dusty air, a 



Fia. 1. 



powerful beam of light, and breathing through a glass 

 tube (the tube actually employed was a lamp-glass, ren- 

 dered warm in a flame to prevent condensation of the 

 breath) across the focus, a diminution of the scattered 

 light was first obserA'ed. But towards the end of the 

 expiration the white track of the beam was broken by a 

 perfectly black gap, the blackness being due to the 

 total absence from the expired air of any matter com- 

 petent to scatter light. The experimental arrangement 

 is represented in fig. 1, where g represents the heated 



' Proc. Eoy. Inst. vol. vi. p. 9. 



