M. Koidzumi 
24:] 
and several minute points of structure, which can hardly be made out in 
fixed and stained preparations, are clearly observable in fresh specimens. 
Dark-ground illumination and vital staining methods were tried and proved 
very fruitful in many cases. For the fixative, Schaudinn’s solution was almost 
exclusively used, and the smears, prepared on cover-glasses, were fixed by 
dropping them face downwards on the fixative. The most excellent and 
frequently used stains were iron-haematoxylin methods. I used in most cases 
2 per cent, solution of iron-alum and ripe alcoholic solution of haematoxylin. 
Giemsa’s fluid, thionin, fuchsin, and other aniline dyes were also used and 
found useful in several cases. Some bacteriological stains were also used for 
the study of special structures. Preparations fixed and cut into sections were 
also made, but were found unsuitable for morphological study; and my 
experience taught me that this method is apt to lead to erroneous judgments. 
Thus total preparations were chiefly relied upon, except for the purpose of 
studying the position of the protozoa in the intestine and their relation to 
its wall. 
3. DESCRIPTION OF FORMS FOUND IN JAPANESE TERMITES. 
The present paper embodies the results of my studies of forms referred 
to eight genera. Six of these eight are genera known already to occur ex¬ 
clusively in the intestines of termites; the other two are new, and belong to 
the same group. Other protozoa were also found in the termites studied. 
These forms are of smaller size, and while some are trichomonad types, others 
are of peculiar organization. My studies of these forms are, however, not com¬ 
plete, so that I shall refrain in this paper from making any further mention 
of them. 
Before entering into a description of each form, a brief synopsis of the 
genera and species will be given here, as I think it will be helpful to the 
reader. 
Synopsis of Genera and Species. 
Our forms are conveniently divided into two main groups, which may be 
called (1) the Trichonympha series and (2) the Pyrsonympha series. These 
groups are not to be taken as constituting families or higher orders: they are 
used here merely for convenience in description. The classification of the 
organisms contained in them will be discussed later (p. 300). The first series 
contains form§ provided with a very large number of flagella, and the second 
those with only four or eight flagellar cords. The species belonging to the 
former series are found in both Coptotermes and Leucotermes, while those of 
the latter are limited to Leucotermes. 
(1) Trichonympha series. 
Six genera are placed in this series. It may be subdivided into twd groups, 
of which Irichonympha and Holomastigotoides may be regarded as the repre¬ 
sentative types. 
16—2 
