M. Koidzumi 
207 
to that where the granule is situated, and embedded in the ground substance, 
which now has a distinctly oval shape (PI. XII, fig. 30). The thread connecting 
the daughter nucleus with the daughter nipple soon disappears, and reorganiza¬ 
tion of the nuclear structure and the reformation of the nuclear sac are carried on. 
In PI. XII, fig. 31, is shown a young nucleus undergoing reorganization. Here we 
see the ground substance of the nucleus becoming reniform, with the chromatic 
threads embedded in it, and a distinct clear zone surrounding the whole mass. 
The latter structure is peculiar to this form. The boundary between the clear 
zone and the endoplasm is distinct, indicating clearly the existence of a 
membrane, and one deeply stainable granule is found on its anterior border. 
As regards the nature of the above-mentioned structures, I believe that the 
reniform mass represents the nucleus in process of reorganization, and the 
membrane at the outer boundary represents the new nuclear sac. Unfortu¬ 
nately I could not completely trace the formation of the new nuclear sac; 
but I believe that the neck and flange-like part are developed from the above- 
mentioned membranous wall, and hand in hand with the process of reorgani¬ 
zation, the clear zone under the membrane becomes reduced, until it finally 
almost disappears around the nucleus. Some further discussions of these 
points will be given in Part II. 
(2) Teratonympha mirabilis var. formosana var. nov. 
(Plate XII, figs. 24 and 25). 
As already remarked, the form of this genus found in the termites of 
Formosa differs from that just described, occurring in Japan proper, and was 
regarded as a distinct variety. The variety formosana is distinguishable from 
the type species by the following characters (see PI. XII, fig. 25). The neck of 
the nuclear sac is shorter, being usually half as long. Further, the main body of 
the sac enveloping the nucleus is found in the preceding form applied closely 
to the nuclear membrane, so that the sac is not usually clearly distinguishable 
from the nuclear wall; but in this variety it is separated from the nuclear 
membrane, leaving a distinct space between them. Both these differences 
are clearly visible in the living animals, as well as in stained preparations. 
Dividing forms are rather frequently met with in this variety, and the 
process of division is identical with that observed in the preceding form. 
IV. Microspironympha. 
This is a new genus proposed for a small form harboured by the Formosan 
species of Leucotermes. Some of the forms taken by Leidy (1881) and Porter 
(1897) to represent possible young stages of Trichonympha agilis resemble 
this form closely, but owing to the insufficiency of the descriptions of the 
American authors, the identification of their forms is hardly possible. As 
compared with Trichonympha , our form is remarkably simple in organization, 
and differs so distinctly that the existence of a genetic connexion is hardly 
