th^ 



280 



THE BEGINNINGS OF LIFE. 



The 



pellicle that forms at first is^ however^ not 



always persistent : after twenty-four or thirty-six hours 

 it may sink to the bottom^ whilst another gradually 

 takes its place which may prove more durable. It is 

 not very plain why some pellicles break up' and sink 

 in this way, but it would seem very probable that such 

 an occurrence may be associated with an imperfect 

 secretion or formation of that transparent jelly which, 

 in ordinary cases, so much helps to give it coherency 

 and strength, and whose presence is probably as ne- 

 cessary in order that subsequent evolutional changes 



In some infusions or fermentable solutions, 



may ensue. 



however, no distinct pellicle is ever formed, Flocculi 

 may appear in the clouded^liquid, which, after a time, 

 sink to the bottom of the vessel ^ or, without the 

 formation of flocculi, a deposit gradually accumulates, 

 whilst the previously clouded supernatant liquid be- 

 comes more or less clear. 



Occasionally it happens that the substance of a 

 pellicle may be almost wholly composed of minute 

 Torula cells — Bacteria being well-nigh absent. I once 

 saw a very re;markable instance of this in an infusion 

 of turnip. In certain of the cases, also, in which no 

 distinct pellicle forms, the fine sediments or floccuU 

 which gradually collect at the bottom of the vessel 

 more especially when the infusion has an acid reaction- 

 are found to consist either^ wholly or largely of vegetating 



Torula cells. 



r 



■ ^ In two or three cases I have failed, after a long search, to find a 

 single Bacterium amongst the myiiads of Torula cells. 



of m'''^^ 



as a g^r 



) 



,k.Iy than B^^^^ 



1 



It 



cf a low* 



solution in wl 



is to the orig 



ce may 



, though 



is purpose than 

 in distill' 



that Ba:t 



•^ch solution 

 N and merely 



w 



1 



1) 



\ 

 I 



ar 



e 



. ^'Diersi 



our hours 

 if the watd 

 stage of 

 surface of 

 ion len32 



t 



the 



Diemb 



er, 



K[ lo 



A 



I 



on 



, ■ O PC 



N, 



cell 



i 



