ENTAMOEBA HISTOLYTICA 4 1 



in unlimited numbers.* After various trials, I found that dividing 

 organisms could only be studied satisfactorily in sections. The infected 

 kitten must be killed, and not allowed to die ; and the ulcers in its 

 intestine must be fixed immediately after it has been killed. The least 

 delay in removing the tissues causes the disappearance of dividing 

 organisms or the appearance of abnormal division forms. Scraping the 

 amoebae out of the freshly excised ulcers also gave me unsatisfactory 

 results as a rule. The following account is based, therefore, upon a. 

 study of the dividing amoebae found in the ulcers of kittens experi- 

 mentally infected with E. histolytica, and studied in serial sections of 

 material fixed and stained by various reliable methods. I have now 

 studied a large amount of material from kittens — obtained during the 

 investigations undertaken with Dr. H. H. Dale in I9i6f — and I have 

 been able to study the process of division in considerable detail. It may 

 be noted, however, that it is by no means so easy to obtain all the stages 

 of division as one might suppose. One may section many ulcers without 

 being rewarded by finding a single stage. 



The process of division in E. histolytica is closely similar to that 

 of E. rananim, and the division of the nucleus is almost identical with 

 that which I have previously described in the cysts of the latter species 

 (Dobell, 1909). I have illustrated the chief stages in E. histolytica in 

 tigs. 43-54, PI. in, which almost describe themselves. 



The first recognizable stage in division (fig. 43) shows an increase in 

 the volume of the nucleus, a fragmentation of the karyosome, and 

 apparently a migration of chromatin from the nuclear membrane to- 

 wards the centre. Such stages are easily recognized, and are connected 

 by numerous transitional stages with organisms showing ordinary resting 

 nuclei {e.g., fig. i, PI. I). The chromatin granules gradually become 

 more numerous, the outline of the nucleus becomes oval, and achromatic 

 threads appear within it — usually more or less longitudinally disposed 

 (fig. 44). I have been unable to make out any definite arrangement of the 

 chromatin granules at these stages : and although the nuclear membrane- 

 stains readily now, and at all subsequent stages, it seems to be no longer 

 studded with chromatin granules on its inner surface — as in the resting 

 nucleus. The appearances suggest that these granules have passed, for 

 the most part, towards the centre, where they form part of the mass 

 of granules now seen (fig. 44). 



The nucleus next becomes fusiform, at first having the shape of a. 

 small stumpy spindle (fig. 45), and later of a long slender one (fig. 46) 

 which ultimately stretches right across the organism (fig. 47). I have 

 studied all these stages with great care, but they are very puzzling. At 

 first sight the spindles suggest mitotic figures, with chromosomes and 

 achromatic spindle-fibres. More careful investigation has always con- 

 vinced me, however, that the irregular masses and threads within the 

 nuclei cannot be resolved into the definite chromosomes and other 

 structures of a true mitotic figure. Fig. 45, which is drawn from a 

 rather small individual at an early stage, will illustrate my meaning. 

 The granules and threads within this nucleus strongly suggest a mitosis. 



* Contrary to Swellengrebel and Schiess (1917), and some other observers, I find 

 that the amoebae in the cat are morphologically identical with those m man. Cf. Dale 

 and Dobell (1917). 



t Vide Dale and Dobell (191 7). 



