THE EVOLUTIONARY RELATIONSHIPS OF BACTERIA I49 



Bacteriaceae, especially from the Acrobactcr group [c.f. Bissct, 1952). It is, 

 in addition, contrary to all experience in other natural groups to assume, even 

 if the cocci were so simple in torm as they appear to be, superficially, that 

 they should, for that reason alone, be considered to be primitive, hi almost 

 all parallel cases, such apparently simple forms have proved to be descended 

 from more complex creatures, and this is almost certainly true of the cocci. 



The scheme of Kluyvcr and van Niel is open to criticism upon other 

 grounds. It postulates four, divergent lines of evolution from the " ancestral " 

 coccus ; the peritrichously flagellate rods, culminating in the sporing bacilli ; 

 the cocci, culminating in sporing sarcinae (a group of very doubtful antecedents, 

 since some at least are environmental variants of sporing bacilli) ; the polar 

 flagellate rods leading to spirilla ; and fourthly, the progressively more 

 branched forms, without flagella. This hypothesis postulates a multiple, 

 independent origin for spores, flagella and the rod form, and even for Gram- 

 positivity ; unless it is assumed that, having all these divergent potentialities 

 in an immediately available genetic condition, this simple coccus was descended 

 from a flagellated, rod-shaped, sporing ancestor. And this, although for 

 entirely ditFerent reasons, 1 consider it probably was ! In its present form, 

 however, the scheme is illogical and must be rejected. At the same time, it 

 provides no clue to the relationships of bacteria with other groups, but appears 

 to assume a separate origin from a tiny, spherical proto-cell, whereas their 

 possession of flagella suggests a close relationship between bacteria and those 

 flagellate protista which are the probable ancestors of all living cells, except 

 blue-green algae. So far as the latter are concerned, no further comment is 

 required upon the classical dictum that they share with the bacteria the lack 

 of a true nucleus. 



Fig, 66 {See p. 148) 

 RELATIONSHIP OF COCCI AND BACILLI 



Preparations by Hale's method of the cell walls of Bacillus, compared with those of Gram- 

 positive cocci. The former (1, 2) are symmetrically subdivided by cross-walls. The smaller 

 cocci (3, 4) are normally divided by a single cross-wall ; the larger cocci (5, 6, 7) are also 

 symmetrically subdivided into four or more cells, but each cross-wall is produced at right- 

 angles to the preceding, a, complete subdivision ; b, an earlier stage of subdivision. 



