STUDIES ON AVIAN ILEMOPROTOZOA. 7J7 



blood (figs. 4, 28) ; but frequently the proximal, rhizoplastic 

 portion of the flagellum is not well stained, and then the 

 blepharoplast cannot be made out. In preparations of culturjil 

 forms it is generally conspicuous, and now and again very 

 prominent (figs. 10, 71, 81, etc.). 



As regards the details of commencing division, the try- 

 panosornes in the blood have provided me, as already stated, 

 with hardly any indications at all. On the other hand, 1 

 have obtained a nice series of stages among the cultural 

 parasites. The first act in the process is apparently the 

 division of the blepharoplast at the base of the flagellum 

 (fig. 120). This is followed by the splitting of the flagellum 

 for some distance, which may be fairly short or fairly long 

 (fi_gs. 100, 104, 121, and 123) ; the splitting never extends, 

 however, throughout the whole of the attached part of the 

 flagellum. In the case of this avian parasite, the splitting-off 

 of a portion of the old flagellum to form the foundation of the 

 new daughter one appears to be of general occuri-ence. I 

 have observed nothing which would indicate that the daughter- 

 flagellum is formed as an entirely independent outgrowth 

 from the second blepharoplast. Fig. 89 shows a flagellum 

 caught in the act of dividing, the proximal portion being 

 drawn out transversely, as a broad band, prior to splitting. 

 In figs. 88, 100, 104, and 123, the newly formed part is still 

 connected at its tip with the old flagellum ; and in fig. 121 

 the new portion,, in this instance only short, has just separated. 

 Of course, once the rudiment, as it were, is cut off, its further 

 growth is quite independent. 



The division of th^ nuclei may begin while the splitting of 

 the flagellum is proceeding (figs. 104, 123), or it may be 

 delayed until the latter process is completed (figs. 90, 121) ; 

 there is apparently considerable variation in this respect. 

 The first stage in the division of the trophonucleus is most 

 probably the division ot" the intra-nuclear centrosome, which 

 acts as a division-centre; this is clearly shown in fig. 88. 

 All that can be said from Gieinsa-stained preparations as to 

 the rest of the process is that the nuclear substance becomes 



