STUDIES ON AVIAN H^MOPEOTOZOA. 699 



of the parasite, certain of whicli appear distinctly unhealthy. 

 The authors state that all the fip^ures on the plates to 

 which I am now referring (as well as others) are of para- 

 sites from a culture in the seventh generation, grown for 

 seven days, by which I understand them to mean a sixth 

 subculture, itself of seven days' age. This long-continued 

 cultivation doubtless accounts both for the varieties of form 

 present, as well as for the number of clusters. Their fig. 2, 

 PI. 9, supposed to represent an early stage in rosette- 

 formation, shows a large indefinite-shaped parasite, in which 

 irregular multiplication of the nuclei is going on. There is 

 no indication of the development of any flagella, and I have 

 no hesitation in regfardingf this individual as an abnormal, 

 degenerating form. That it would ever give rise to a rosette 

 of active, flagellate parasites is most improbable. Again, 

 fig. 3, PI. 9, represents an agglomeration cluster of four or 

 five somewhat similar forms, three or four of which, however, 

 are not quite so degenerate, as they still possess flagella; but 

 the same irregular multiplication of the nuclei is shown. 



Phases such as these have, I venture to say, no connection 

 whatever with the rosettes of more typical parasites figured 

 on PI. 8. Fig. 2, here, is a small cluster of a dozen pyriform 

 individuals, each with a single, centrally directed flagellum. 

 Not one of the individuals shows the least sign of division. 

 Similarly in fig. 1, PI. 8, there is a cluster of about eighteen 

 parasites. Hence, in neither of these rosettes is there any 

 evidence that they are going to give rise to one of many 

 more individuals, such as that of fig. 4, PI. 8, by multiplica- 

 tion. And, from my own experience, I know that such 

 rosettes can be formed very quickly indeed. In this cluster 

 of fig. 4 there are several individuals at the periphery, which 

 are m^anifestly only loosely attached, and whose flagella 

 cannot be connected with the central core (cf. my own 

 figures). There can be no doubt that these are the individuals 

 which have been most recently attracted to the cluster. 



A point in favour of this view of Novy andMcNeal's would 

 be furnished by evidence which went to show that two 



