220 KE. A. MINCHIN. 
somes with flagellum morphologically anterior, considers 
that the Herpetomonas ancestor was “a primitively intes- 
tinal or entero-ccelomic! parasite of an invertebrate,” not 
necessarily of a blood-sucking one, in which their entire life- 
cycle took place; that when the invertebrate became blood- 
sucking in habit, its gut parasites found themselves in a - 
nutrient medium in which they were able to multiply 
enormously and were thus prepared for life in vertebrate 
blood, into which they finally succeeded in passing by 
inoculation through the proboscis. Their primitive habitat is 
supposed to be shown by the fact that only non-sexual multi- 
plication takes place in the vertebrate, and the invertebrate 
host is necessary for their sexual development—a statement. 
which, it may be remarked, is by no means definitely proved, 
however probable, for trypanosomes. 
Léger’s theory is logical and complete ; it only seems to me 
to present one flaw. If we consider the transmission of 
trypanosomes generally, we find that it does not always take 
place by the intermediary of an insect, but may be effected 
by a leech; and if we include Heemosporidia, as Léger does 
in his theory, we then have to reckon with Arachnida as well, 
in some cases. In other words, the constant trypanosome host 
is the vertebrate; the inconstant host is the invertebrate. I 
propose, therefore, to consider the facts from the standpoint 
of an opposite hypothesis, namely, that the ancestors of. 
trypanosomes were primitively parasites of the gut of verte- 
brates, like so many flagellates known to exist at present, and 
that from the gut they passed into the blood of the vertebrate 
and finally into the gut of the blood-sucking invertebrate. If 
we attempt to imagine and to reconstruct on this basis the 
successive stages in the evolution of trypanosome life-cycles, 
we should probably have the following series of events : 
(1) The ancestral form, ex hypothesi, was a flagel- 
late parasitic in the vertebrate gut, which doubtless was 
1 T am not guite sure that I understand the meaning of this adjective, 
though it has a familiar sound to the morphologist. 
