788 HEMOPHILUS 



and longer and somewhat stouter rods may predominate. Other strains, again, 

 may present an entirely different picture, the bacilli being thin, long, wavy or curved 

 and sometimes lying together in tangled masses. In films prepared from strains 

 showing any of these diverse morphological types, but especially those which 

 show some proportion of thread forms, it is not unusual to encounter large, spherical 

 swollen bodies, often attaining a diameter of 2-3 /bt, or even more. These are some- 

 times attached to the end of a thread, sometimes laterally, and sometimes appar- 

 ently at the end of a short lateral stalk (Wade and Manalang 1920, Kristensen 1922). 

 Another form which is occasionally met with consists of a long thread with an 

 enormous fusiform swelling, situated centrally or towards one end. 



This morphological diversity raises a problem of considerable difficulty from 

 the point of view of classification. Any one of the types we have referred to may 

 predominate in a single strain ; and two strains may yield a microscopical picture 

 so different that it is difficult to believe that we are dealing with a single bacterial 





/ 



,if 



Fig. 163. — H. influenzcs. Fig. 164. — H. influenzoe. 



From 24-hours' culture on Fildes' agar, From 24-hours' culture on Fildes' agar, 

 showing typical cocco-bacillary forms showing short and long bacillary forms 



(X 1000). (X 1000). 



species. The only test of the real significance of such morphological differences is 

 their constancy ; and this test is not so easy to apply as might at first sight appear. 

 It is quite certain that many strains maintain their morphological individuality 

 over long periods of artificial cultivation, involving numerous successive sub- 

 cultures ; and Dible (1924) regards these morphological characters as sufficiently 

 well differentiated, and sufficiently stable, to justify the recognition of several 

 different varieties, or types. Many of those who have studied this group, on the 

 other hand, have rejected morphological criteria as a basis of classification within 

 the group, on the following grounds. If a large series of strains is carefully examined 

 the variation in form is not found to be discontinuous. In a small sample of strains 

 it is easy to obtain an appearance of discontinuity ; but, unless one classes all 

 strains which depart from the typical cocco-bacillary form in a single heterogeneous 

 group, it is not possible to define a limited number of morphological categories, 

 to which all strains can be definitely assigned. Between the minute, short bacilli, 

 and the long tangled threads, there exists a long range of intermediate forms. More- 

 over, while some strains are morphologically homogeneous, others display, in the 



