ENTAMOEBA HISTOLYTICA 43 



events depicted. For my own part, I do not consider the nuclear 

 division of E. hislolytica to be a regular mitosis. On the other hand, 

 1 cannot call it an amitosis. It seems rather to belong to an intermediate 

 category, like the nuclear division of E. rananun. 



A word may be added regarding the possible presence of a centriole. 

 It will be noted that I have not so far described a centriole — for the 

 reason that I have not found one. In the early stages — e.g:, that of 

 fig. 43 — I have never succeeded, by any method of staining, in detecting 

 a centriole. In later stages, such as figs. 44 and 45, " centrioles " can, 

 no doubt, be discovered by the ingenious. Their selection is not 

 difficult, to the willing, from the number of granules and threads at 

 one's disposal. Of the existence of a real centriole, however, I am 

 unable to convince myself. I have made a special study of all the 

 spindle figures such as those shown in figs. 45-47 : because it seemed 

 to me that if centrioles were present they should here be discoverable 

 at the poles. But I have not found them. Occasionally — as in 

 fig. 47 — a granule may be visible at one pole, or more rarely at both. 

 Sometimes, also, threads appear to join these granules, forming a 

 " centrodesmose." But I attach little importance to these, as they may 

 be found in different positions in different nuclei ; and if they represent 

 centrioles or their derivatives, then these structures must display a degree 

 of mutability and variation in behaviour which makes their relation to 

 nuclear division open to the gravest doubts. In short, I have found no 

 structures which I can regard, with any confidence, as centrioles or 

 centrosomes in E. histolytica ; and in this respect my findings agree 

 completely with those previously recorded for E. ranarum (Dobell, 

 1909, 1914). 



The division of E. histolytica has been partially described by several 

 workers, but nobody appears to have studied all the stages previously. 

 Schaudinn (1903) stated that the organism divides into two, and that 

 the nuclear division is an amitosis ; but he did not describe the process. 

 Most of the published figures depict organisms in various stages of 

 degeneration — not in division. For example, the "dividing" forms of 

 Werner (1908), and Hartmann's (1908, 1912 a) "prophases" with 

 "dividing centrioles," are probably not division stages at all. It is 

 significant that Hartmann never found the later stages of division. 

 Brumpt (1913, p. 24, fig. q) has figured some division stages in amoebae 

 from the cat, but these also are incomplete and partly abnormal. The 

 nuclear division is not completed in the way suggested by his figures. 

 Job and Hirtzmann (1916) say that the nucleus divides by "amitosis," 

 but give no description or figures. It seems probable that they never 

 observed real dividing forms. Mathis and Mercier (1916 b) have given 

 a brief description of division in E. histolytica, but it is incomplete and 

 faulty, as they apparently saw only a very few stages in human stools. 

 I cannot confirm their statement that a centriole is present ; but until 

 they publish a fuller account, with figures, of the stages which they 

 actually saw — apart from their interpretations — it is impossible to discuss 

 satisfactorily the discrepancies in their description. It should be noted, 

 however, that they deny that the " histolytica forms " of £. histolytica — 

 the large tissue-inhabiting forms, which contain red corpuscles— undergo 

 division at all. These, of course, are the very forms which do divide, 

 and which constitute the bulk of the species. They are the forms whose 

 divisions I have just described in detail. 



