72 
C. H. MARTIN AND MURIEL ROBERTSON. 
flagellum, with the old membrane flagellum, persisting in its 
connection with the other moiety of the blepharoplast. The 
new free flagella., one in the case of the former individual 
and two in the case of the latter, are probably formed by 
splitting from the edge of the membrane flagellum (compare 
PI. 12, fig. 31). This splitting process, as will be seen later, 
is also met with in the passage from the resting to the active 
stage. 
We are inclined to regard the relatively rare stages found 
with more than three free flagella, and with no other sign of 
division, as examples in which the flagellar apparatus has 
undergone precocious development in the transition towards 
the dividing form (PI. 12, fig. 35). 
One of the most characteristic features in the division of 
Trichomonas eberthi is furnished by the behaviour of the 
chromatic blocks. These at an early stage show a tendency 
to become arranged in a single line parallel to, and apparently 
in connection with, the axostyle. During the early stages 
of division there is a marked tendency of the chromatin in 
the nucleus to become condensed into a number of masses. 
In early stages the number of these masses se£ms to be eight 
(PI. 12, fig. 40), but in later stages there is a tendency for 
them to become paired, giving rise to four double masses 
(fig. 41). Unfortunately we have no figure showing the 
actual state of the nucleus at the time of division, but it is 
probable that the latter is a very modified form of mitosis. 
We are also unable to supply any really conclusive evidence 
as to the behaviour of the axostyle during division, but an 
inspection of a large number of stages leads us to believe 
that the old axostyle undergoes some process of solution, 
starting from its anterior end, followed by a re-formation in 
the case of both individuals along the lines of the chromatic 
blocks (cf . PI. 12, figs. 41, 42, 43). This hypothesis may seem 
at first sight rather improbable to those who regard the 
axostyle as a relatively permanent skeletal structure, but we 
have so often watched the development of the axostyle in 
resting forms of Tri cho rn astix gallinarum within acorn- 
