TRYPANOSOMES IN TSETSE-FLIES AND OTHER DIPTERA. 191 
tomonas-like forms, in which 7 is invariably in front of N.) 
The forms with » posterior or terminal are interesting, as it 
is very probable that they represent very nearly the form 
under which 'l’. grayi occurs in the vertebrate host; and in 
this connection their tendency to preponderate in the red 
blood is significant. I recognise now as erroneous my former 
statement (P. R.8., B 78, p. 243) to the effect that the forms 
with posterior were nuper parentes, though this may 
be so in a few cases (see p. 192). It may be said, however, 
that for the forms occurring in the intestine of the fly, 
the normal and typical position of 1 is a little in front of N, 
for not only is this the condition found in those forms (the 
slender form and the Herpetomonas form presently to be 
described) in which the position does not vary, but it is also 
the position which n always assumes when division is about 
to commence. 
I have made a detailed study of the process of division in 
T. grayi, with the object. of being able to distinguish clearly 
between the stages of multiplication and of conjugation. 
The process of division shows some variations; I will first 
describe what appears to be the most usual course of events, 
and after that I will deal with the deviations from this course 
that are met with. 
The first event in the process of division is that the minute 
basal granule or blepharoplast (see p. 172) of the flagellum 
divides into two (fig. 151). At the same time the kineto- 
nucleus becomes slightly enlarged and elongated. Starting 
from the two blepharoplasts the flagellum splits longitudinally 
and as the two minute blepharoplasts are at first connected 
by a delicate thread, we obtain a common and characteristic 
early stage of division, in which, in a trypanosome otherwise 
of normal aspect, the flagellum shows the appearance of a 
loop at its proximal end (figs. 204, 205). The splitting of 
the flagellum continues and the two blepharoplasts travel 
away from each other, so that the appearance of a loop is lost; 
but at this stage there is nearly always seen a crossing over 
of the two halves of the flagellum in the region where it is 
VOL. 02, PART 2,—NEW SERIES. 15 
