MORPHOLOGY 181 



and those, like the bacillus of whooping cough, for which the growth factors are 

 not essential as nutrients. 



Ducrey's bacillus in any event is heemophilic on the narrower definition, but 

 it is doubtful whether the Morax-Axenfeld bacillus would be included in this 

 genus by any such redefinition, and we have therefore described it in Chapter 37, 

 together with others that cannot at present be assigned to any named bacterial 

 genus. The Bordet-Gengou bacillus cannot, we think, be dealt with in the same 

 way. It resembles H. influenzcB so closely in morphology, in habitat, and in many 

 other ways, that it would certainly be placed in close association with it by any 

 systematic definition that did not rely exclusively on a narrow nutritional criterion. 



In 1911, Ferry in the United States described a short bacillus that he had 

 isolated from the respiratory tract of dogs in the early stage of distemper ; to 

 this organism he gave the name of Bacillus hronchicanis. A similar organism 

 had been described as early as 1896 by Galli-Valerio, by Tartakowsky (1897-98) 

 in 1898, by Strada and Traina in 1900 as B. pneumonice caviarum, by Martini 

 in 1900 as B. pulmonum glutinosus, and by Selter in 1906 as B. cavisepticus mohilis ; 

 Ferry, however, was the first to study it fully. Later Ferry (1912, 1912-13) 

 found the same bacillus in a guinea-pig epizootic, and in monkeys and rabbits ; 

 he therefore changed its name to B. bronchisepticus. This organism has in recent 

 years been included in the Brucella group, whose type species, Br. melitensis, it 

 resembles in both individual and colonial morphology, and in its inability to 

 ferment carbohydrates. However, the conspicuous degree of antigenic similarity 

 between B. bronchisepticus and H. pertussis that has become evident in the past 

 seven years demands a reconsideration of its classification. Antigenic relation- 

 ships alone do not establish a taxonomic relationship, but in this case other resem- 

 blances are sufficiently good to warrant provisional inclusion of B. bronchisepticus 

 in the group containing H. pertussis. Thus, B. bronchisepticus produces in guinea- 

 pigs lesions similar to those produced by H. pertussis in rabbits and puppies (Smith 

 1913, Mallory and Hornor 1912, Mallory, Hornor and Henderson 1912). Both 

 H. pertussis and B. bronchisepticus are natural pathogens in the upper respiratory 

 tract, and, according to Rhea (1915) B. bronchisepticus lesions in rabbit lungs 

 are similar to the lung lesions of human whooping cough. And, finally, the 

 general cultural resemblances between H. pertussis and B. bronchisepticus are as 

 close as those between B. bronchisepticus and Br. melitensis (Chapter 34). We 

 have therefore provisionally included both the Bordet-Gengou bacillus and 

 B. bronchisepticus in the genus, and have emended the generic definition accord- 

 ingly (see Lwoff 1939). It should, however, be pointed out that H. bronchisepticus 

 grows very much more readily than any other member of the Hcemophilus group — ' 

 so much so that it has been classified by some workers with the paracolon bacilli. 

 Such a difference, taken together with the free growth on plain nutrient agar and 

 the possession of motility, renders the classification of this organism in the 

 Hcemophilus group necessarily tentative. 



Morphology. — H. injluenzce, as originally described by Pfeiffer (1893), and as 

 most commonly seen in strains recently isolated from cases of influenza, is a short 

 rod, so short as to be almost coccal. It is very small, 1-1-5 /j, by 0-3-0-4 /u, with 

 rounded, sometimes rather pointed ends. In some cultures these cocco-bacillary 

 forms are the only forms seen. More usually, among these predominating short 

 forms, are found a proportion of longer bacilli, and a few long thread forms. In 

 other cultures, the cocco-bacilli may be relatively scanty, or altogether absent ; 



