N. H. SWELLENGRBBEL 
119 
By subsequent division, large rosettes of preflagellate forms may be 
produced, such as Patton describes. 
Preflagellate stages may persist for a long time, not only in the 
crop, but also in the midgut (Diagram VIII, Figs. 1-3); they possess 
a short flagellum and move about actively. In some flies the midgut 
only contains preflagellates; in such cases, the latter might be 
mistaken for an independent species. The preflagellates show a 
peculiar invagination of the ectoplasm, just where the flagellum 
becomes free; this invagination takes a pinkish hue with Giemsa, 
proving that it is of ectoplasmic origin; it forms a short canal through 
which runs the flagellum. I suppose that this invagination may be 
regarded as a rudimentary cytostome, such as may be found in a 
complete state in free-living flagellates. 
The preflagellate forms develop into real herpetomonads. Fig. 4 
shows a transitional stage in this part of the life-cycle. 
3. Post-flagellate stages. 
When the post-flagellate stages are formed in the rectum, the 
flagellate becomes smaller (Diagram IX, Fig. 1); the blepharoplast 
I. Normal flagellate becoming smaller with short flagellum. 
2-5. Elimination of the achromatic part of the blepharoplast. 
6-7. Last stages before the production of the “ cysts."’ 
8-10. “Cysts” without a cyst-'s^all; Fig. 9 shows a “cyst” with double I’hizoplast. 
II. “ Cyst ” with a “ cyst-wall.” 
