286 
Intestinal Protozoa in Termites 
in the large individuals it is frequently vacuolated, and great numbers of 
globular bodies of a peculiar kind are commonly found in it. These globules 
are rather refractive in fresh specimens, and stain with eosin somewhat deeply, 
but faintly with iron-haematoxylin 1 . As in many members of the preceding 
group, food debris is usually present in the protoplasm, but consists not only 
of vegetable matters (wood, etc.) but also of the bodies of other animalcules, 
such as small individuals of Dinenym'pha. Animalcules ingested into the 
endoplasm are found rounded and shrunk into spherical masses, and various 
stages of their disintegration, probably due to digestion, can be recognized. 
Dividing forms were met with very rarely in this species, and I was barely 
able to make out the chief stages of the process, after an eager search through 
large numbers of termites in both Japan proper and Formosa. Division seems 
to take place invariably in the forms living freely in the lumen of the gut. 
In individuals going to divide, the body becomes rounded and the axial 
filament detached from both the nucleus and the body wall and set free in 
the endoplasm. The nucleus becomes oval or round in shape, and probably 
separated from the body wall, except at its anterior pole, so that it remains 
attached at only this one point, from which the flagellar cords arise. The 
nuclear network now disappears, the karyosome breaks up, and globular masses 
of chromatin make their appearance scattered in the almost structureless 
nuclear sap (PL XIV, fig. 58). Thick rod-shaped chromosomes are now formed 
from the chromatin, and a mitotic figure, somewhat different from that seen 
in the trichonymphids but resembling rather closely that of metazoan cells, 
makes its appearance. No structure corresponding to the strand seen in all 
forms of the preceding series is visible; but a distinct achromatic spindle is 
formed, with the chromosomes arranged at its equator (PI. XIV, fig. 60). The 
chromosomes divide transversely, and the daughter groups move towards 
opposite poles in due course. The nuclear membrane persists throughout 
the entire process, and minute centrioles can be made out at the poles of the 
spindle in the later stages. 
The flagellar cords remain connected with the nucleus, four of them being 
found attached to each pole of the spindle (PL XIV, fig. 61). Once I noticed 
a short fibrous strand on the membrane of the nucleus in an early stage of 
division (Pl. XIV, fig. 59). Unfortunately I have been unable to ascertain its 
origin and fate, or to determine its relation to the ends of the cords; but I 
am inclined to believe that it represents a stage in the formation of the 
achromatic spindle. 
The axial filament gradually degenerates in the endoplasm and disappears 
before the division of the nucleus is completed. Shortly after the daughter 
nuclei are formed, two new axial filaments can be seen, one attached to each 
of them (Text-fig. D). As regards the origin of the new axial filaments I have 
1 The vacuoles are usually present in only the posterior two-thirds of the body. It was 
the presence of such vacuoles which apparently led Comes (1910) to bestow the specific name 
vacuolata upon the Pyrsonympha which he regarded as belonging to a new genus Lophophora. 
