A. Porter 357 
The two organisms so formed may separate at once or the division may 
be repeated, a rosette of parasites then resulting (Fig. 20). 
As there is great variation in the size of the cysts ingested, 
there is also much variation in the size of the pre-flagellates exhibiting 
division. 
Longitudinal division of mature flagellates. I have frequently 
watched the entire process of longitudinal division of the fully flagellated 
living forms. In many cases, the forms about to divide were slightly 
stouter than their neighbours, but this was not invariably so. 
Division always seemed to be initiated by the blepharoplast becoming 
dumb-bell shaped and gradually dividing into two (PI. IV, Fig. 48). 
This was followed by vigorous movements of the flagellum which began 
to split into two (Fig. 49) at the end near the blepharoplast. The 
nucleus also divided into two, the daughter nuclei migrating outwards 
to the periphery of the body (Fig. 49). The flagellum then executed 
rapid, lashing movements, and the split extended to the body (Figs. 50, 
51, 52). The undulating membrane divided with the flagellum. 
After that, the progress of division was more gradual. The daughter 
organisms diverged, a Y being formed by their free portions with the 
undivided part (Figs. 53—56). Ultimately an angle of 180° was 
attained (Fig. 57), and the final separation occurred. Each daughter 
form thus produced had half of the nucleus, blepharoplast, membrane 
and flagellum of the parent. As a rule, division is equal (Figs. 54, 55) 
though, unlike Patton, I have sometimes seen markedly unequal division 
(Fig. 56). Possibly this is peculiar to the Crithidia of G. paludum. 
By repeated longitudinal division rosettes of parasites are formed in 
the gut, but these very rapidly break up. In a true division rosette 
(Fig. 20), the flagella lie towards the periphery of the group, whereas 
in the more common aggregation rosettes (Fig. 47), the flagella serve 
as a centre of entanglement or are applied to the food particles to 
which the Crithidia cling, the posterior ends of the bodies of the 
parasites being outwardly directed. 
Division of very narrow flagellates (Fig. 58) results in the formation 
of parasites that are almost spirochaetiform in appearance (Figs. 59, 60). 
Polymorphism in C. gerridis, resulting from division. 
The multiplication by division early in life and also at a more 
mature stage of existence is sufficient to account for the very great 
variation in size seen in the Crithidia of Gerris paludum. Some forms 
Parasitology ii 
23 
