1 



r^ 



282 



THE BEGINNINGS OF LIFE 



i 



In these several patches there may be seen delicate 

 ovoid Tcrula cells of almost any size beneath 



6000 



form, and we see only 



50000 



in 



diameter. The larger cells are united in little groups 

 of twos and threes, and budding from them may be 

 seen pullulating projections of different sizes. Separate 

 cells also exist, smaller and smaller in size, till at last 

 they cease to be cellular in 

 peculiarly refractive dots or specks less than 

 diameter. In other places a colony of Torula cells 

 seems to be about to grow up. Here there may be 

 seen merely one or two of the smallest bodies which 

 distinctly display the cellular form interspersed amongst 

 a variable number of the refractive specks of all sizes 

 down to the mtnlmtim ^vhtble stage. And when such a 

 patch is marked and watched at different intervals 

 a crop of perfect Torula cells is soon seen to occupy 

 the same situation. The Tcrula cells do undoubtedly 

 multiply pretty rapidly by a process of gemmation^, 

 when they have attained their full size, and possibly 

 also they may increase by processes of fission durin 

 their earlier stages. Accordingly, their distribution is 

 such as might have been expected amongst such self- 

 multiplying units. Very rapid processes of sub-division 

 cannot be recognized amongst ordinary plastide-par- 

 ticles and Bacteria^ although many persons assume that 

 such phenomena do take place 2, and, moreover, when 



^ M. Pouchet doubts the occurrence of this mode of multiplication 

 (*Nouvelles Experiences/ &c., 1864, p. 168). 



2 I have actually seen the fissiparous division of a Bacterium only 

 on comparatively few occasions. ■ 





,4 '''''' 



.St. As to 



mk cells, tw 



ire 



)( 



the develop 

 .visible, pari 



or 



filam 



are the represei 



«f living mattei 

 The former 

 Halher, of Jen; 

 »f briefly ep 

 toming plastic 

 'micrococci.' 



''^'^s of plasm 

 H^i to such 

 ^' believes tl 

 ision of 



the 



breaki 

 Wcr 



in 



, ''^^ not 



the 



PrO( 



i 



) 



\ 



tob 



yof 



e 



tl 



