90 C. CLIFFORD DOBELM. 
chromatin is suspended in the form of granules of variable 
size. 
The first phase of division is characterised by the nucleus 
becoming elongated and somewhat fusiform (fig. 35). The 
central body appears to throw off granules of chromatin at 
opposite poles. <A little later the fusiform shape gives place 
to a more or less oblong form, like a short, blunt rod (fig. 36). 
Enlargement of the ends of the rod next takes place, with 
the formation of the characteristic dumb-bell figure (fig. 37). 
The chromatin masses itself at the ends of the dumb-bell, 
the intermediate portion becoming band-like and staining a 
lighter tint (fig. 38). Minute chromatin granules may be 
seen scattered throughout. The two ends of the dumb-bell 
then undergo increased differentiation and separation, be- 
coming rounded off, and containing the aggregated chromatin 
elements of the nucleus. For some time these daughter- 
nuclei remain connected by a fine filament as in fig. 39. 
Finally the filament disappears, and two completely formed 
daughter nuclei are left (fig. 40). 
A general discussion of the nuclear phenomena will be 
found on p. 103. I will here merely call attention to the fact 
that division is amitotic; there are no differentiated chromo- 
somes, no extra-nuclear division centres. In this respect it 
resembles that of some rhizopods (cf. for example, the 
Leydenia-phase of Chlamydophrys). 
(8) The flagellum.—(Pl. 4, figs. 2-5; Pl. 5, figs. 41- 
46). A description of the manner in which the new flagella 
are formed, as seen in the living monad, has already been 
given (p. 88). The finer details of the process can be made 
out only in stained preparations. The flagellar insertion is 
most clearly seen in osmic vapour preparations (fig. 41). 
In these the basal granule is very distinctly seen, lying in 
the wall of the cytopharynx against the reservoir. As we 
have already seen, the flagellum is retracted before division. 
In stained specimens it can further be seen that it gradually 
disappears in the direction of the basal granule, until only 
this structure is left (figs. 42 and 43). The basal granule 
