68 SULPHUR BACTERIA 



are very adaptable. For example, certain bacteria lead both 

 a saprophytic and a parasitic existence according to the nature 

 of the habitat ; and there are denitrifying bacteria that undergo 

 the sulphur metabolism. The insufiticiency of a physiological 

 grouping is apparent in the sulphur bacteria, for whilst the 

 group includes almost every known variety of form and 

 structure, the possession of sulphur inclusions does not connote 

 any other distinction common to the whole group ; and this 

 is the surest indication that the classification on the basis of 

 a common physiological trait is not genetically sound. But 

 it is nevertheless adopted in this book on account of its con- 

 venience in limiting the area of investigation. The further 

 subdivision of the group is followed on morphological lines. 

 All existent classifications have the defects of the classifi- 

 cations of higher plants in the days of Linnaeus. These were 

 based on artificial distinctions rather than on natural afiinities, 

 and to a certain extent the difficulty will always remain in the 

 classification of lowly forms with the high degree of plasticity 

 possessed by bacteria. Of the earlier attempts the best is that 

 by Migula (i — 3), which is based on morphological characters. 

 This classification would have been completely successful were 

 it not for the fact that the distinguishing features selected 

 by him were in several instances transient and not permanent 

 attributes. 



In the classifications of Engler (1912) and of Meyer (1912), 

 the sulphur bacteria are incorporated in a general scheme in 

 which a few changes are made in the division of the sulphur 

 bacteria to bring them into conformity with the general scheme 

 adopted by the authors. There had not been during the 

 period (1909-12) any investigations which threw any further 

 light on the afitinitics of the sulphur bacteria. One addition 

 had been made to their number, namely, Hillhousia, dis- 

 covered by West and Griffiths in 1909. 



Of the recent classifications those of the American Society 

 of Bacteriologists are not completely satisfactory. They 

 violate the first essential of any logical system of grouping 

 which demands that it should not be possible to fit any 

 organism into more than one group. Inclusion in one group 



