Gr. H. F. Nuttall and Gr. S. GtRAham-Smith 245 
exhibit radii, they become rounded and an achromatic zone appears 
round the dark chromatin mass. The number of degenerating forms 
increases after the third day. The process of multiplication in defi- 
brinated and diluted blood could not be observed, and no multiplication 
took place when the developmental forms were transferred to fresh 
normal or diluted blood. 
Kleine thinks that in canine piroplasmosis the parasite usually lies 
upon the red blood corpuscle. We have elsewhere stated our reasons 
(■ Journ. Hygiene, vi. 636; vn. 250) for regarding these parasites as 
being intracorpuscular, and see no reason for changing our opinion 1 . 
Kleine also considers that the cultivation forms with radii are usually 
epicorpuscular, but here again we hold a contrary opinion. Owing to 
its extreme delicacy and lack of staining power the corpuscular envelope 
may perhaps have escaped his notice in many instances. Occasionally 
the corpuscular membrane cannot be demonstrated and then the parasite 
appears to be free. 
The writers experiments. 
In our attempts to cultivate P. canis under artificial conditions we 
have made use of several different methods and media, which are 
described in the following pages. Under most of these conditions the 
parasites rapidly degenerated without exhibiting any signs of multi¬ 
plication, or changes of form suggesting further development. These 
negative experiments (A—L) are first described. 
In the last series of experiments (M) however certain large forms 
with long radiating processes were encountered, similar to those de¬ 
scribed by Kleine and Koch. 
A. In order to ascertain whether the parasites would enter and 
multiply in normal dogs’ corpuscles kept outside the body under 
artificial conditions the following experiments were made. Blood¬ 
stained serum taken shortly after death from the heart of a dog dead of 
piroplasmosis was centrifugalised, and drawn up into capillary pipettes. 
1 Kleine’s statement and that of Kinosbita (1907, p. 309) regarding the epicorpuscular 
position of Piroplasma canis appears to be based upon deceptive appearances in stained 
preparations. Kleine’s paper is illustrated by excellent coloured figures and microphoto¬ 
graphs 'which show appearances essentially similar to those we have observed. We have 
not however seen large agglomerations of radiate bodies such as he figures on Plate IV, 
fig. 14, and Plate V, fig. 12. 
Kinoshita apparently confined his attention to the study of stained preparations of 
P. canis. He has consequently advanced various hypotheses regarding the development 
of the parasite which we do not consider to be justified. 
16—2 
