STUDIES ON PAEASmC PROTOZOA. 
463 
fig. 16) there certainly appeared to be a group of granules at 
each end of the spindle, but this was probably chromatin 
that had travelled there from the equatorial plate, or else the 
residue of the peripheral chromatin.^ 
Gradually the spindle lengthens, and the chromatin from 
the middle separates towards the two poles, there forming- 
dense, dark-staining, club-shaped masses. The spindle extends 
right across the cyst, curved against the bulge of the vacuole. 
It thins out and becomes vague, then disappears, leaving the 
two nuclei. These are at fu st exceedingly small and compact 
(PI. 31, fig. 20), but they presently swell out, and take on 
the characters of the normal flagellate nucleus, with peripheral 
chromatin and a voluminous karyosome. As a rule the 
karyosome is not central, but lies pressed against the nuclear 
membrane( Pis. 31 and 32, figs. 22-25). The daughter-nuclei 
sometimes lie close together, sometimes at opposite sides of 
the cyst, which now is inclined to lose its regular circular 
outline (PI. 32, fig. 26). 
The rliizostyle has meanwhile divided into two. Sometimes 
this division takes place very early, even before that of the 
nucleus, sometimes later. This division does not seem to be 
a splitting of the whole structure ; division of the basal 
granule is followed by the gradual growth of a second rhizo- 
style from one of the daughter-granules (PI. 31, figs. 13, 23 
and 24). 
^ If the recent “ systematisation ” of primitive mitosis given by 
Alexeielf (1913) is to be followed, then the mode of unclear division in 
the cysts of Rhizomastix gracilis comes near to that named 
mesomitosis. “ Centrioles anx deux poles dn noyau en division; la 
plaque equatoriale, formee i^ar le caryosome, massive an debut, se 
morcelle en un certain nombre de chromosomes qui . . . se 
repartissent aux deux poles oii ils se confondent avec les centrioles.” 
But I have been unable to demonstrate centrioles in Rhizomastix, 
and in this respect Alexeielf’ s rheomitosis, in which the centrioles 
are early hidden by accumnlating chromatin granules, seems to meet 
the case better — “ a la place de centrioles on observe des corps chroma- 
tiqnes de jdIus en plus volumineux et finalement on se tronve en presence 
d’une apparence de promitose avec ses corps polaires.” 
