CONDITIONS FAYOFRING FERMENTATIOlsr. 17 



was found that tbe urine became lighter-coloured and turbid in 

 two or three days. Other experiments showed that a slight excess 

 of liquor potassse tended to retard or even prevent the occurrence 

 of fermentation, though a quantity of liquor potassse notably 

 below that needed for neutralization was found to be e£B.cacious 

 in inducing it, and that, too, almost as rapidly as if the neutrali- 

 zation had been complete. Even when the liquor potassse was 

 added in quantity only sufficient for half-neutralization, fermen- 

 tation still took place in many instances, though in such cases the 

 result was usually delayed for five or six days. 



In all these trials it was found that the fluid, when turbid, 

 was not foetid ; its odour was for the most part scarcely at all 

 altered, though at times it was rather more marked than usual. 

 The organisms found in the fermenting urine were in all cases 

 the same, viz. Bacilli, either short, medium size, or in the form 

 of long threads — and not the ferment thought by Pasteur to be 

 the invariable cause of the conversion of urea into ammonic car- 

 bonate and water*. Sometimes only the short unjointed rods were 

 found, though more frequently these were mixed with varying 

 amounts of longer Vibrio-like bodies, and with threads such as I 

 and others have generally spoken of as Leptothrix. 



The results of the foregoing preliminary experiments induced 

 me to seek other, stricter methods, free from two possible sources 

 of fallacy which might be thought to have influenced the results. 

 Thus, as the whole of the tube containing the liquor potassse was 

 not immersed in the boiling fluid, it was possible that the heated 

 vapour within the- tube was not certainly sufiicient to sterilize 

 the small quantity of air also contained within it above the level 

 of the liquor potassse. It would have been easy to meet this 

 source of uncertainty by boiling the closed liquor-potassse tubes 

 in a vessel of water for a time before inserting them into the expe- 

 rimental flasks containing the urine. But the other possible 

 source of fallacy would still have remained. It might be said by 

 some that the cotton-wool plug, which hitherto had been deemed to 

 be thoroughly efficacious as a protective barrier between the im- 

 purities of the outside air and the boiled fluid, was itself a nidus 

 for germs, some of which, unkilled by the steam of the boiled fluid 

 (by which, of course, the wool has been saturated), subsequently 

 found their way into the fluid within the flask. This objection has 

 been urged by Prof. Tyndall against some experiments made by 

 Dr. William Eoberts ; and if it is a valid objection (which I very 

 * Aun. de Ohim. et de Phys. t. Ixiv. (1862) p. 50, 



LINN. JOUEN. ZOOLOGY, YOL. XIY. 2 



