246 THE HAEMOFLAGELLATES 



M'Neal) have maintained the view that the Invertebrate is the 

 primary host in all cases. Minchin, however, has always considered 

 the Vertebrate as the principal host ; and in his latest memoir 

 on the Trypanosomes of Tsetse-flies (58), proofs of which he very 

 kindly allowed the writer to see, he regards all Trypanosomes as 

 descended from an intestinal Vertebrate form, and indicates the 

 lines upon which the evolution may be supposed to have advanced. 

 This ancestral form produced resistant cysts for dispersal, and thus 

 contaminative infection was brought about. (It would be extremely 

 interesting to ascertain whether the intestinal Trypanoplasmata 

 known (see p. 249) have such a cyst-formation.) The next stage in 

 evolution is when the parasite has penetrated the intestinal wall, 

 and come into relation with the circulatory system. Until it came 

 into relation with a blood-sucking Insect, this type would have to 

 pass back into the alimentary canal for dissemination. So far, we 

 have no evidence of an existing instance of this stage. Subsequently, 

 the blood-parasite became adapted to an Insectan host, in the gut 

 of which it encysted, reinfection of the Vertebrate being by the 

 contaminative method. T. grayi in all probability furnishes an 

 example of this type. Lastly, the parasite is thoroughly adapted 

 to the biology of the Insect and passes forwards to the front part 

 of the alimentary canal : infection of the Vertebrate is now by the 

 inoculative method. This may possibly be combined in some cases 

 with the contaminative mode, but probably in most encystment no 

 longer takes place, being unnecessary (e.g. the lethal Trypanosomes, 

 Piscine forms, etc.). 



Of course, in those cases where, as above remarked, the Verte- 

 brate is probably the secondary host, there is no reason to suppose 

 that, as a rule, the parasites leave the circulatory system. 



Phylogeny. As stated at the beginning of this article, the 

 Trypanosomes, as a whole, are to be regarded as including two 

 entirely distinct families, in one of which (the Monadine type) the 

 attached flagelium becomes free at the true anterior end, and in the 

 other (the Heteromastigine type) at the true posterior end. The 

 former type is derived by the progressive migration backwards of the 

 kinetonucleus towards the posterior (aflagellar) end, in the manner 

 well illustrated by Leger's series of Herpetomonadine forms (see 

 Fig. 29). The latter type is derivable from a Trypanoplasmatine 

 ancestor itself in turn doubtless to be derived from a Bodo-like 

 form by the loss of the anterior free flagelium ; * so that the non- 

 flagellate extremity is the true anterior one. 



The writer is unable, owing to limits of space, to enter fully 



1 A comparison of the different degree of development of the flagella in various 

 forms is instructive as illustrating the manner in which the Trypanoplasmatine condi- 

 tion may have resulted from that found in Bodo, and its further evolution. 



