Hirudinea. 
i6i 
A couiparisoii of a large series of specimens from Japan and China in my own 
collection shows that the present species is somewhat variable in this respect, the 
extremes being represented by the usual Japanese form with only one longtitudinal 
band, and the Chinese form with five longitudinal bands of almost equal breadth 
which I have provisionally named var. sinica in my collection. The present specimen 
belongs to the latter group. A specimen from near Tateyama Bay, in the possession 
of the British Museum and figured by Blanchard in his paper on Asiatic leeches 
{1896), exhibits a condition just intermediate, the lateral bands being here sufficiently 
distinct but markedly narrower than the median one. The dark spots on the ventral 
surface are also variable, being much less numerous in the present specimen. than in 
most of the Japanese examples. 
3. Myxobdella' annandalei, n.g.,n.sp. 
Plate VII, figs. 1—5. 
There are in the collection three specimens of this very interesting leech. It is 
allied in many respects to the well-known genus Haemopis {Aulastoma) , but differs 
from this, as well as from all other genera, in having the somites distinctly bounded 
externally. This is certainly a rare exception among the Hirudinea, in which the 
metameric structure is, as a rule, completely obliterated by the formation of 
secondary furrows looking exactly like those separating the somites. As is well 
known, the fixing of somite-limits has long been the subject of debate and an artifi- 
cial method of plotting out the somites by assuming the segmental sensillae to be on 
the first annulus was universally adopted until as late as 1900, when Castle's 
important paper on the metamerism of the Hirudinea appeared, in which the author 
pointed out the inadequacy of this procedure. Were the present form known at the 
time of Gratiolet (1862), there would have been obviously no discussion about, and 
no false determination of, the somite-boundaries in this group of annelids. 
The study of this leech was greatly facilitated by a number of additional examples 
sent to me by Prof. H. E. Earle, of Hong Kong University, who collected them at 
my request at the same locality as Dr. Annandale found his specimens. I wish here 
to express my cordial thanks to Prof. Earle for his kindness in obtaining for me the 
valuable material. 
Shape and Dimensions. All three of the specimens are a good deal contracted. 
They are of nearly the same size, the length measured along the curved dorsum being 
26-29 mm. and the greatest width about 9 mm. The shape of the body is much like 
that of a Haemopis in a contracted state, only with the head end somewhat narrower. 
Both extremities are bent downwards, especially the posterior, so that the hinder 
sucker is entirely hidden when the animal is viewed from above. Toward the anterior 
end the body tapers gradually and ends in the anterior lip of the small oral sucker, 
which is not marked off from the adjoining region. The dorsal surface is convex 
throughout, while the ventral side is slightly concave in the greater part of the length. 
1 liv^a mucus, SSeWa leech. 
