TfilOHOMASTIX SERPENTIS. 453 



basal grannie. This latter loses its connection with the 

 nncleus, and divides, so that two bnnches of fiagella are 

 formed. It has been fonnd impossible to connt the flagella 

 at this stage, or to determine how the new ones arise. Con- 

 comitantly the nucleus becomes dumb-bell shaped [fig. 4]. 

 The organism then assumes a roughly triangular appearance, 

 but is rapidly drawn out, and becomes somewhat sausage 

 shaped. A constriction appears in the middle [fig. 6], and 

 it is now possible to make out that there are four flagella at 

 either end, one in each case directed away from the other 

 three. The nucleus is still dumb-bell shaped, and situated 

 mesially. The daughter-cells now rapidly draw apart [fig. 7], 

 and remain connected by only a small bridge of protoplasm, 

 at either end of which is a nucleus, the daughter nuclei 

 being still connected by a very fine protoplasmic strand. 

 Further separation now takes place, finally resulting- in the 

 snapping of the connecting strand of protoplasm [figs. 8, 9]. 

 In the daughter individuals which are thus separated, it will 

 be seen that the nuclei lie at the extreme posterior end of 

 the body, and there is no axial rod. Up to this stage the 

 process usually occupies about twenty minutes. If one of 

 these daughter cells be carefully watched, it will be seen that 

 the nucleus travels forwards, and as it does so the axial rod is 

 seen to be developed in the region immediately posterior to 

 it. The rod appears to be formed by the nucleus in its track 

 as it passes forward [figs. 9, 10]. After some time — usually 

 one to two hours — the nucleus reaches the anterior end of the 

 body, and enters once more into relation with the basal 

 granule of the flagella. In some way the axial rod becomes 

 connected with the basal granule, but it is impossible to see 

 how this is effected. The cytostome appears to be lost prior 

 to division, and to be re-formed in each of the daughter cells, 

 but I am not quite clear on this point. 



This account of the division of T. serpontis differs from 

 that of T. lacertge as described by Prowazek. If I under- 

 stand this author correctly, the axial rod would appear to 

 function as a kind of division centre ; for it retires towards 



VOL. 51, PART 3. NEW SERIES. 34 



