APPENDIX. 327 



tory. After their introdnction the liquids were boiled for 

 five minutes in an oil-bath. 



The first experiment in this shed resulted in complete 

 failnre. Not one of the infusions exposed to the moteless 

 air of the shed escaped putrefaction. 



Either of two causes, or both of them combined, might, 

 from my point of view, have produced this result. First, 

 a flue from the laboratory was in free communication with 

 the atmosphere not far from the shed ; secondly, and this 

 was the real cause of the infection, my assistants, in pre- 

 paring the infusions, had freely passed from the laboratory 

 to the shed. They had thus incautiously carried the con- 

 tagium by a mode of transfer known to every physician. 



The infected shed was disinfected ; the infusions were 

 again prepared, and care was taken, by the use of proper 

 clothes, to avoid the former causes of contamination. The 

 result was similar to that obtained at Kew, viz. organic 

 liquids which, in the laboratory, withstood two hundred 

 minutes' boiling, were rendered permanently barren by 

 five minutes' boiling in the shed. 



A third clear issue is thus placed before us which I 

 should hardly think of formulating were it not for the 

 incredible confusion which apparently besets this subject 

 in the public mind. A rod thirty feet in length would 

 stretch from the infusions in the shed to the same in- 

 fusions in the laboratory. At one end of this rod the 

 infusions were sterilized by five minutes' boiling, at the 

 other end they withstood two hundred minutes' boiling. 

 As before, the choice rests between two inferences : — 

 Either we infer that at one end of the rod animal and 

 vegetable infusions possess a generative power which at 

 the other end they do not possess, or we are driven to 

 the conclusion that at the one end of the rod we have 

 infective and at the other end uninfective air. 



The second inference is that which will be accepted by 

 the scientific mind. To what, then, is the inferred differ- 

 ence at the two ends of the rod to be ascribed ? In one 

 obvious particular the laboratory this year difiiered from 



