TRICHOMONAD FLAGELLATES. 353 



"giant forms divided abnormally commonly giving rise to three or 

 four daughter cells." He further states that division, in living forms 

 under observation, rarely became complete, a conclusion which our 

 observations indicate was merely the result of the slowness of the pro- 

 cess and its metabolic phases. That it may be regularly completed 

 is indicated in our text figure F, of Trichomonas augusta. He further 

 concludes that the whole finally fuses into a large amoeboid mass 

 which finally dies. This may be true in moribund cultures but we 

 do not regard this as the normal outcome of the process. 



The occurrence of multiple mitosis in the Hexamitidae is definitely 

 established by the observations of Noc (1909) and of Prowazek and 

 Werner (1914, pi. 10, fig. 13) on Laviblia intestinalis in which they 

 find and figure individuals with eight nuclei, that is, a trophozoite 

 which has undergone two of the three synchronous mitosis, for it is 

 obvious that the plasmodial stage of the binucleate Lamhlia corre- 

 sponding to the 8-nucleate plasmodium of Trichomonas will have six- 

 teen miclei. Noc (1909) also figures two multinucleate structures 

 with an indeterminate number of nuclei, which he regards as multiple 

 division stages of Lamhlia, though the morphological basis for such 

 reference requires further elaboration. 



Our observations demonstrate the occurrence of multiple mitosis 

 in normal fresh material from normal hosts in Trichomonas augusta 

 Alexeieff, T. muris Hartmann, Eutrichomonas serpentis Dobell, and 

 Tetratrichomonas proivazeki Alexeieff, as well as in certain other 

 trichomonads an account of which will be given in later papers. The 

 process is thus one of widespread occurrence and seemingly a wholly 

 normal one in the Tetramitldae {Trichomonas, Eutrichomastix, Tetra- 

 trichomonas) in which it results in free-moving non-encysted plasmodia 

 composed of eight merozoites which subsequently slowly detach them- 

 selves singly from the plasmodium. In some of the Hexamitidae 

 {I^amhlia) on the other hand, the developing plasmodium comes to 

 be more or less encysted (Prowazek and Werner, 1914) apparently 

 as a rule, and the method of release of merozoites is not as yet evident. 

 Plasmodium formation by multiple fission is thus a normal process of 

 general occurrence throughout the Polymastigina as defined by Doflein 

 (1913). 



The repeatedly offered inference that the multinucleated, multi- 

 flagellated Trichonymphida are related to the Polymastigina there- 

 fore gains added probability, for the temporary multinucleated, multi- 

 flagellated Plasmodium of the Polymastigina, or more especially of 

 the Hexamitidae, may be regarded as an evolutionary step in the 



