THE INTESTINAL FLAGELLATES OF INSECTS 251 



species of flies a different species with trypanosome-like forms 

 which could not be thus interpreted. Likewise this process of en- 

 cystment and the pecuHar structure of the cyst as described by the 

 writer for H. muscce-domcsticcc has not been confirmed in all 

 respects. Are all rounded forms true cysts? Here are suggestions 

 for further study. 



The time has come for a general reinvestigation of the flagellates 

 of insects with the problem of specificity in mind. It has been 

 shown experimentally that one species of protozoon may establish 

 itself naturally in the digestive tract of more than one host species. 

 Conversely, in a number of cases it has been shown that one host 

 species may harbor two or more species of one genus of intestinal 

 protozoa. Experimental work has thus borne out the "principles" 

 as set forth by Alexeieft*. A number of problems present them- 

 selves. 



The general problem of host-specificity, or the specificity of 

 any particular flagellate for its insect host may be considered first. 

 There is nothing of the nature of an unexplainable determinism 

 about host-specificity. It may be defined as the adaptation of one 

 species (the parasite, in the broad sense) to a complex of physical 

 and chemical factors which obtain in (or on) another species or 

 a more or less limited group of species (the host, or hosts). If the 

 number of available host species is one, we say the parasite shows 

 rigid host-specificity ; if considerably restricted in suitable hosts we 

 say that the specificity is limited ; if the number of suitable hosts 

 is comparatively large, or if they are fairly distantly separated 

 taxonomically we speak of loose host-specificity. There is no para- 

 site, however, which is not greatly limited in available hosts. 



The writer (1923) and Drbohlav (1925) have shown that 

 Herpetomonas viiisccc-domesticce can produce sustained infections 

 in a number of genera and species of muscoid flies. This proves, 

 apparently, that this flagellate may have more than one natural 

 host, and that we cannot always give a parasite a new specific name 

 simply because it is found in a host where it has not been previously 

 noted. There are other herpetomonads and crithidias which need 

 a similar reinvestigation, for example, the parasites in the siphon- 

 APTERA in which group Wenyon (1926) lists sixteen species from 

 which flagellates have been described. At least twelve different 

 specific names are employed for these flagellates. This whole group 

 deserves a reinvestigation involving a c}1:ological study of the 



