203 
tlie parasites may be found in the blood in various forms at all periods 
of infectioa Their development does not appear to occur in 
successive phases related to alternating presence and absence of 
rnpanosomes in the bl(X)d, as is the case with T. gavibicnse and 
I. equiperium. The elucidation of the developmental relationship 
of the various forms which thus exist together in the blood would at 
first sight seem to present a certain amount of difficulty, but in 
reality this difficulty is not so great as it appears. Thus the various 
authors who have already considered the subject are fairly well 
agreed with regard to the relationship in a developmental sense of the 
various forms one with another. Medium-sized parasites, such as 
those represented in fig. i, certainly give rise by growtli to the large 
pointed types represented in figs. 1 1 • 18, so also these latter 
unquestionably pass into the still larger round and regular multi- 
nucleated masses, such as those represented in fig. 23. .Such masses 
may again in turn be found in all stages of breaking up into smaller 
bodies, and this process of dissociation certainly produces the 
characteristic rosettes, and other forms of temporary association 
commonly met with (figs. 2G-30). Longitudinal division among 
the medium-sized forms such as those represented in fig. i, appears 
to be a rare occurrence, but that it does take place is indicated by such 
types as those represented in figs. 2-5, wherein the nucleus or the 
extra-nuclear centrosome, or botli, have become divided, and a 
second flagellum has arisen. We have, however, ourselves not 
encountered the late phases of division in such forms of T. lewisi, and 
have consequently only been able to figure the early stages of the 
process. The forms such as those represented in fig. i appear 
to certainly arise from, and to merge into, forms such as those 
represented in figs. 37-38, and these undoubtedly in turn have arisen 
trom the products of the dissociation of the large multi-nucleated 
^nasses. Thus we appear to have a cycle of development in the blood, 
"hich, starting from any particular type such as those represented in 
b passes through the phases represented in figs. 12-ig into 
the large forms represented in figs. 24-25. From this stage the cycle 
continues through stages such as those represented in figs. 30-38, 
finally through growth and division the individual derivatives of 
wulti-nucleated masses pass back to the formation of types such as 
those originally chosen, as the starting point, and represented in fig. i- 
