270 Herpeto monads 
on other characters “ d’acquisition ancienne et paraissant actuellement 
soustraits a l’influence du milieu.” As such they select the number of 
the flagella, and the presence or absence of a rhizoplast 1 2 . 
My observations on the herpetomonads in dung-flies are in accordance 
with those of Patton. The formation of the second flagellum is the 
immediate forerunner of longitudinal division: in the flagellate as it 
appears in the larva this is still more evident. The bi-flagellate 
appearance simply depends on the time at which the longitudinal split 
occurs,—in the parasite in the larva it is, as I have said, usual to find 
splitting at a very early stage, and bi-flagellates are therefore rare. 
The cause of this different behaviour in the fly and in the larva is 
obscure, but I regard it as an expression of the effect of a slightly 
different medium on the same organism. 
Then as to the second “ constant ” Herpetomonas character—the 
presence of a long rhizoplast: I have showed that the flagellates in the 
fly are usually provided with a long, well-marked rhizoplast (the 
thickness, however, depending a good deal on the stain employed), 
while in the flagellates from the larva the root of the flagellum is very 
much shorter, and is often no thicker than the flagellum itself 3 . In 
certain Giemsa preparations the flagellar root does not stain definitely, 
hut appears as a diffuse pinkish area: such cases seem comparable to 
Ckatton’s description and figures of Leptomonas agilis, but may merely 
be due to faulty staining (fig. 28). 
If the classification of Chatton and Alilaire be accepted, then the 
only logical conclusion is that I am here dealing with two different 
parasites—a Herpetomonas from the fly, and a Leptomonas from the 
larva. I prefer to regard the parasite as the same throughout, and 
I consider that the mere fact that the rhizoplast may assume such 
different appearances under different conditions, does away with its 
value in classification. 
1 The authors do not explain how those characters are useful with regard to Crithidia, 
which is the genus particularly offensive to their logical sense. 
2 For my part, I am inclined to regard thickening of the flagellum root, i.e. “a well- 
marked rhizoplast,” as the beginning of another division. Supposing that the splitting of 
the flagellum root is the first step in longitudinal fission, it may often happen that a second 
division is preparing before the first is completed. 
The inconstancy of the rhizoplast is also indicated by the fact that two such careful 
observers as Berliner and Porter working independently on H. jaculum in 1909 are entirely 
disagreed as to whether a rhizoplast is present in this form or not. 
