928 PROTOZOA AND OTHER ANIMALS 



scribed the early infection of young nymphs of Kalotertnes flavicollis by 

 direct application of the mouth parts, accompanied by sucking, to the 

 end of the abdomen of the dealates (Fig. 195). 



Certain small polymastigotes are often retained through the molting 

 period (Kirby, 1930; Child, 1934). In Zootermopsis this is true of the 

 minute forms Trkercomitus and Hexamastix. At the last molt the situa- 

 tion differs from that in the preceding molts. Child (MS) reported 

 that in the last molt of Zootermopsis, although the number of flagellates 

 is greatly reduced, all species are carried through from the seventh 

 instar nymph to the winged imago. Cross (MS) and May (MS) have 

 confirmed this fact in Kalotermes minor and Zootermopsis; the shed 

 intima of the nymph, still containing Protozoa, is retained within the 

 gut; and the Protozoa later escape into the lumen of the imago's intestine. 



In Cryptocercus the Protozoa are not lost at the time of molting; but 

 then, and only then, most of them form either well-defined cysts (in 

 Trichonympha and Macros pironympha) or resistant stages (Cleveland, 

 et al., 1934). Flagellates are present, often in great numbers, in pellets 

 passed in the first few days after ecdysis. Cleveland found that some 

 pellets, passed immediately after molting, consisted mostly of Protozoa 

 in encysted or resistant form. Reinfection of defaunated roaches took 

 place when they were placed with molting roaches; but not, except occa- 

 sionally with smaller polymastigotes, by association with other infected 

 roaches. Proctodaeal feeding, then, does not have the same role in trans- 

 mission in Cryptocercus as in termites. Cleveland could not find out the 

 exact manner in which infection is first acquired, but thought it probable 

 that it is by association with molting individuals. Once acquired, the 

 faunule persists until the death of the roach. 



It is probable that cross infection has not been a significant factor in 

 determining the present distribution of flagellates in termites, below the 

 Termitidae at least. Furthermore, the unique characteristics of almost all 

 the flagellates, which have no close relatives except in roaches, indicate 

 that the faunules do not to any great extent include acquisitions from 

 other arthropods or other animals. The present distribution of the flagel- 

 lates, the absence of resistant stages, and the isolated habits of termites 

 support this opinion. If it is sustained by further studies the flagellates 

 of termites will be shown to be easily the leading group of animals for 

 correlative studies in phylogeny of symbionts and their hosts. 



