174 THE FLOATING-MATTEE OF THE AIR. 



while where the closure was but slightly defective we 

 had the escape of a fraction. 



The defects thus revealed came, I concluded, into 

 play when the infusions were introduced, the descending 

 column of liquid sucking in minute air -bubbles between 

 the india-rubber tubing and the pipette, thus carrying 

 with it the external contagium. Few seem aware of the 

 precautions which are sometimes essential to save the 

 experimenter from error in inquiries of this nature. 

 Even with some of our best and most celebrated observers 

 I find no adequate sense of the danger involved in their 

 modes of experimentation. 



§ 14. Experiments in the Royal Gardens, Keiu. 



But it was only in exceptional instances, dependent 

 on the state of the air, that even precautions such as 

 those described in the foregoing section secured freedom 

 from contamination. The contagium seemed omni- 

 present and persistent, and whether it was local or 

 general — due to the accidental condition of our labora- 

 tory, or to an epidemic of the air — became a question 

 with me, not by any means to be decided offhand. On 

 this point, then, I held judgment in suspense. The in- 

 fection was, to all appearance, ftdly accounted for by 

 reference to the conditions imder which I worked ; but 

 as regards outbreaks of epidemics the autumn had been 

 a remarkable one, and it seemed well worth investigating 

 whether it was not also a period prolific generally in the 

 germs of putrefaction. 



I resolved therefore to break away wholly from the 

 Eoyal Institution, and, thanks to the friendly permission 

 of Sir Josepli Hooker, I was enabled to transfer my 

 apparatus to Kew Gardens. By the enlightened mu- 



