239 
few large forms of Piroplasma cants, which simulate the crescents of 
aestivo-autumnal malaria, having the chromatin sometimes concen¬ 
trated in the middle, sometimes forming a loose mesh work. These 
parasites were I0'4 to 107 m long and r4 to 17 m broad. They 
regarded them at first as gametes, but in their last paper they do not 
consider them to have any connection with Piroplasma canis. Their 
chief reason was the fact that they only found seven ‘ gametoid 
bodies’ altogether, and these occurred in one animal. No flagella 
could be observed. 
Kinoshita figures somewhat similar parasites (figs. 47, 48, 49) seen 
in the heart-blood, pancreas and lung, some hours after death. He 
refers to his figs. 47 anci, 48 as conjugation forms, and to fig. 49 as 
an ookinet (?) in accordance with Schaudinn’s ideas. 
We have been able to trace the development of large biflagellate 
forms from the normal intra-cellular parasite. Tn the films, where 
large biflagellate forms occurred, along with ordinary intra-cellular 
parasites, forms were also found in which both nuclei were 
considerably enlarged, as represented in figs. 54. 55 - These bodies 
increase in size, and the smaller nucleus in the meanwhile divides 
(fig. 56), often remaining connected with the large one by fine 
chromatic lines. The subsequent changes vary in details, but on the 
whole two main forms of development may be followed. On the one 
hand, an irregular number of small round chromatin masses originate 
from the large nucleus, frequently remaining connected with it by fine 
chromatic lines, which "eventually disappear (figs. 72, 73). From the 
fact that these masses often appear double, it seems possible that they 
divide (figs. 75, 76). At the same time the appearance of the 
large nucleus changes ; the chromatin becomes aggregated at the 
centre, and a lightly staining area is left between it and the well- 
defined nuclear membrane (figs. 60, 72-74). Eventually, two flagella 
are formed, each of which may end in the neighbourhood of a small 
chromatic mass, but in some cases the flagella appear to have no 
definite origin (figs. 74, 77, 78). 
A second mode of development takes a somewhat different line. 
The large nucleus frequently buds off at first a small number of 
granules, and eventually it seems to throw out the whole of its 
chromatin in form of a large densely staining mass ; figures 62-66 
representing different stages of this process. The remainder of the 
