40 RESEARCH IN PROTOZOOLOGY 



tribution of flagellates among termites are based on records, many 

 unpublished, from sixty-six species of termites of the three lower 

 families, forty-four of which the writer has been able to examine. 



1. In general, the faunas of more primitive termites are not of 

 more primitive character. Highly developed hypermastigote flagel- 

 lates have been found in all subfamilies of xylophagous isoptera 

 which have been investigated. Nevertheless the faunas of many 

 KALOTERMiTiD^ are predominantly, and some entirely, polymasti- 

 gote, while those of the rhinotermitid^ are chiefly, and in most 

 cases entirely, hypermastigote. 



2. Trichomonas and related genera and the trichonymphid.e 

 are distributed through all the large groups of xylophagous ter- 

 mites, though they do not occur in every genus and species. Most 

 genera of these flagellates are restricted, so far as is known, to 

 certain families, subfamilies or genera of termites. 



3. The greatest amount of flagellate specialization appears to 

 have taken place in the subfamily kalotermitin^. Of forty well 

 defined genera of flagellates, half are known exclusively from 

 these termites and only ten are not known to be represented in 

 them. At the same time it should be remembered that of the 382 

 species of termites belonging to the three lower families, 215 be- 

 long to the subfamily kalotermitin^. trichomonadid^, deves— 

 coviNiD^, oxYMONADiD^ and CALONYMPHiD^ are very character- 

 istic components of their faunas. Trichonympha is of frequent 

 occurrence. Genera of the jceniid^, Staurojocnina and Spirotri- 

 chonympha are sometimes encountered, while Kofoidia and Hop- 

 lonyinpha are known only from one termite, Ncotcrnies simplici- 

 cornis of Arizona and desert California. 



4. With a few exceptions, every species of the kalotermitin^ 

 which has been examined has a very characteristic flagellate fauna. 

 It is usually possible to identify the termite by examination of 

 this, or at least to distinguish between known species of a region. 

 This provides a means of identifying nymphs, which in the ab- 

 sence of other casts might be confused. In most other groups of 

 termites, the species cannot so easily be distinguished by this 

 method. The faunas of the three species of Tcrmopsis and of two 

 species of Porotcnnes closely resemble each other. Among those of 

 Reticulitcrmcs the dififerences are slight and those of all other 

 RHiNOTERMiTiD^ are much ahke. In the kalotermitin^, the 

 faunas are not at all characteristic of the genera of termites as 



