liFKTKAMIA KIRKMANI. 11 



area of protoplasm clear of granules (fig*. 8, c. a). At this 

 stage the trophozoite may become surrounded by a membrane 

 of very various thickness ; sometimes it is almost or quite 

 imperceptible (fig. 10) ; or it is thin, as in fig. 9 ; or it may 

 form a thick cyst-wall, as in figs. 12 and 13. In the same 

 host and at the same time we can find all these conditions 

 among the different trophozoites. 



In the case of fig. 12 the individual was surrounded by a 

 thick cyst-wall {cij) ; but whether or not this is produced, the 

 mode of formation of the spores is the same (c f. figs. 10 

 and 13). 



In fig. 12 it can be seen that the densely granular tropho- 

 zoite is becoming divided up by trabecular in a manner some- 

 what similar to that which occurs in Sar oocyst is. Ulti- 

 mately the trabecuhe are so formed that each nucleus, 

 together with a certain amount of the cytoplasm, becomes 

 shut off from the remainder (fig. 9). The trabeculee will form 

 the future spore-membrane (fig. 11). In fig. 12 it may be 

 observed that at the time of killing a certain amount of the 

 trophozoite at the tAvo ends {un. p.) had not been invaded 

 by the trabeculas. Such a condition may frequently be 

 observed in Sarcocystis. 



During spore-formation the large granules in the cyto- 

 plasm tend to disappear, and the substance of the trophozoite 

 becomes rather homogeneous in appearance. In the great 

 majority of cases which I have observed the whole substance 

 of the trophozoite is completely divided up into compart- 

 ments, each of which is converted into a single spore. In 

 one or two cases, however, a certain amount of the trophozoite 

 had ]iot become involved in the spore-formation, although 

 such spores as had been formed were completely ripe (fig. 14, 

 r. p.) and had mostly escaped out of the thin cyst. 



These isolated cases may perhaps be regarded as a rever- 

 sion to a more typical Myxosporidium ancestor, for in the 

 great majority of the specimens observed such residual 

 Plasmodium is not present, and it cannot be regarded as 

 normal in this species. 



