TRYPANOSOMES IN TSETSE-FLIBS AND OTHER DIPTERA. 219 
of Léger (24). The second origin is from a Her- 
petomonas-like ancestor, in which the insertion of the 
flagellum becomes shifted backwards to form the undulat- 
ing membrane; Léger’s “trypanosome with the flagellum 
morphologically anterior.’” Woodcock (42) and Liihe (26) 
have gone so far as to base generic distinctions on this 
alleged difference of origin. For evidence bearing upon 
evolution in past times the zoologist turns naturally to 
the documents supplied by the records of development at 
the present time; and it must be acknowledged that so far 
there has been no development evidence whatever forth- 
coming in favour of a Trypanoplasma-like ancestor. No 
one has yet recorded a developmental stage of a trypano- 
some with two flagella, not even in those of fishes, which 
seem most likely to be allied to Trypanoplasma. On 
the other hand, Herpetomonas forms are common, if 
not invariable, in trypanosome development, and we may 
agree with Brumpt (10) and Léger (24) that in such 
cases this form certainly represents the ancestral form of 
these intestinal parasites before they became “ saneuicolous.”’ 
We may therefore look upon the Herpetomonas-form in 
the development of trypanosomes as a true recapitulative 
larval form, a type with which zoologists and embryologists 
are sufficiently familiar in the development of animals of all 
classes. For the present a Trypanoplasma-like ancestor 
most remain hypothetical until concrete evidence for it is 
forthcoming in given cases. 
Léger (24), confinmg his speculations to the trypano- 
1907, p. 707). As I pointed out at the time, this view offers some difficul- 
ties; first, in view of the paleontological facts, showing vertebrates and 
arthropods to have been distinct in Silurian epochs, if not earlier; secondly, 
in view of the fact that ‘‘ Hemoprotozoa” divide their parasitism sometimes 
between an arthropod (insect or arachnid) and a vertebrate, sometimes 
between a leech and a vertebrate. I may point out further that the inverte- 
brate host of a Heemoprotozoon is always one that sucks the blood of the 
vertebrate; it remains, therefore, to be explained how the transmission from 
host to host was effected at the ancestral period when, ex hypothesi, all 
hosts of a given parasite belonged to a single species. 
