312 THE CLASSIFICATION OF BACTERIA 



in applied medical bacteriology, but it is not a sound basis for a classification of 

 bacteria in general. A classification intended to accommodate the species familiar 

 to animal, plant, industrial, biocbemical and " pure " bacteriologists should depend 

 mainly upon characters whose selection is securely based on agreed general prin- 

 ciples. Its design should be consistent in that, should a species be discovered 

 with a hitherto unknown combination of differential characters, it could be accom- 

 modated in the system without dislocating it. Clearly, any working classification 

 will be a compromise between the utilitarian and the " logical " classification based 

 on a priori conceptions of the relations of bacterial species. 



The most promising a 'priori conception upon which to base a classification, 

 and one which has proved fruitful in many other branches of biology, is the con- 

 ception of species in a phylogenetic series. We have already discussed in Chapter 3 

 the relation of nutritional requirements to the possible evolution of bacterial species, 

 and some of the dangers of a too-ready acceptance of the phylogenetic hypothesis. 

 Our chief objection to the phylogenetic conception of the nutritional series was 

 the impossibility of deciding which end of the series — organisms with the most 

 complex or organisms with the least complex nutritional requirements — repre- 

 sented the starting point. With morphological characters there is perhaps less 

 difiiculty, for the spherical shape being the simplest and the most economical 

 shape that could be taken by a unicellular organism (see Thompson 1942) may 

 with some justification be taken as the primitive type. From this primitive 

 coccus we can assume developments in the direction of aggregates of cocci, of 

 bacteria, and of thread and mycelial forms, and in some members of each group 

 postulate the acquisition of flagella, capsules, and other of the more striking 

 morphological features of bacteria. 



In their valuable review of bacterial classification, Kluyver and van Niel (1936) 

 in fact make this assumption the starting point of their proposed natural system 

 of classification. Each of the main morphological groups springing from the 

 cocci is subdivided according firstly to the main sources of energy of the bacteria, 

 and secondly according to the most favoured substrates and their modes of dis- 

 similation. Thus, there are the photosynthesizing autotrophs and heterotrophs, 

 and the chemosynthesizing autotrophs and heterotrophs, four groups that are 

 further divided according to their most favoured modes of dissimilation. 



In choosing these mode.s, Khiyver and van Niel point out that the type of attack 

 is more important than the range of attack. Thus, the difference between an organism 

 which ^lits glucose into lactic acid, and one which splits it into butyric and acetic acid, 

 CO2 and hydrogen, is more fundamental than the difference between two organisms of 

 the first type, one of which attacks maltose. 



The principles have been further developed by Stanier and van Niel (1941), who 

 propose the term Monera to cover all micro-organisms without true nuclei, plastids and 

 sexual reproduction. Fig. 46 summarizes the proposed arrangement of Monera. The 

 reader is referred to the original paper for details of their system, which may be com- 

 pared with that set out on p. 319. 



In our preoccupation with bacteria, parasitic on the larger animals, we are, 

 however, concerned with distinctions that as yet do not fall into the province 

 of the morphologist and biochemist, and while agreeing with the criteria governing 

 the subdivisions that can be achieved in this province, we shall not find them 

 particularly helpful with characters like pathogenicity or antigenic structure. 

 Both are expressions of certain biochemical features in the organism, and may 



