252 
Intestinal Protozoa in Termites 
number of chromatin masses, which are found assembled at the centre, forming 
a large spherical heap surrounded by a clear zone. In the majority of cases, 
the chromatin masses are distinct in their outline and separated from each 
other. Sometimes, however, they are not so sharply contoured and distinctly 
isolated, but loosely connected to form a somewhat reticular mass. The 
chromatin masses in any one nucleus are of nearly uniform size; but in some 
individuals they are small in size and numerous, and in others few but large. 
Despite my long search, carried out through all seasons, I have been able 
to find only a small number of dividing individuals. The observations of 
American authors agree with mine in this point. In the Formosan variety, 
on the other hand, dividing forms are found not infrequently, and I have 
had ample material for the study of the process of division, which seems to 
be identical with that of the variety under consideration. Thus no separate 
account of the process in the variety japonica will be given. 
No forms which suggest the occurrence of other modes of development 
have been observed by myself. Leidy has figured several forms which he 
thought to be the young, but a contrary view was afterwards put forward 
by Porter. Some of the forms considered by Leidy to be the young are really 
met with in Leucotermes speratus in association with the present variety. But 
there is strong evidence indicating that they have nothing to do with Tricho¬ 
nympha, as pointed out by Porter, and have to be classified under different 
genera. The form which was taken by Porter—though with much misgiving 
—to “present the best evidence of being the young' 5 is not found in Leuco¬ 
termes speratus, though it is frequently met with in Leucotermes flaviceps 
accompanied by the other variety of Trichonympha. It is thus quite certain 
that there is no genetic connexion between Porter’s “young form' 5 and the 
variety now under consideration. 
(2) Trichonympha agilis var. formosana var. nov. (Plate X, figs. 5-12). 
This variety is distinguished from the foregoing by its shape, the propor¬ 
tional dimensions of various parts of the body, and by the structure of some 
parts. 
The difference in shape of the two varieties is recognized at a glance 
(compare Figs. 1 and 5, Plate X). The body length is nearly the same in the 
two varieties, while the breadth of formosana measures only a little more 
than one-half that of japonica. The organism assumes a spindle or a lanceolate 
shape, in contradistinction to the oval shape of the former variety, and the 
inclination of the wall of the bell is conspicuously acute. The head (including 
the nipple and the bell) occupies some one-sixth of the length; and it is thus 
a little smaller than that of the variety japonica. Though the body is slender 
in this variety, as described above, the nucleus is about the same size as in 
japonica, so that a larger portion of the breadth of the body is occupied by it 
in this form. Some differences are also recognizable in the nipple and the 
