6o ZOOLOGY OF THE FAR EAST. 
In European lakes the loo-metre line has been accepted by limnologists as a 
convenient but conventional boundary between the shallow-water and the deep-water 
fauna, just as marine zoologists accepted the loo-fathom line. In Lake Biwa no spot 
is deeper than 320 feet ' (about 97 metres) and no consider- 
Limitsof "Deep" and" Shai- ^^^^ ^^^^ ^^^^^ deeper than 75 metres, but in practice, 
low water. 
though the boundary is not exact, we find that the upper 
bathymetric limits of the deep-water fauna lie somewhere very near 100 feet (== a 
little over 310 metres). 
RUPICOLOUS FORMS. 
The shores of Lake Biwa are for the most part flat and either sandy or muddy, 
but towards the northern end of the lake, especially on islands such as Chikubushima, 
there are rocks that descend abruptly into the water to a 
Spcaes from the lower sur ace (^gp^]^ jgg^g^ 20 feet , whilc botli at the samc localities 
of small stones. 
and at the south end, the bottom at certain spots is covered 
close inshore with small stones. In most cases the small stones are of artificial origin, 
but they seem to have encouraged the assemblage of a somewhat characteristic 
fauna, consisting, so far as the molluscs are concerned, of small species of gastropods 
that find a secure hiding place on their lower surface. 
Lithotis japomca, Choanomphalus japonicus (with its variety suhstriatulus) and 
Planorbis hiwaensis are characteristic .species of this fauna. 
Melania niponica and M. biwae do not seek concealment in the same way, but 
crawl openly on vertical as well as horizontal surfaces and appear to attain their 
maximum development only on large rocks. The shells of M. niponica found on 
stones at Otsu are small and much eroded, not exceeding 20 mm. in length and 9'5 
mm. in maximum breadth, and their sculpturing is comparatively ill-developed. 
Shells of the same species found on small stones at Chikubushima are less eroded, but 
only a little larger and in other respects similar ; it is only on the rocky shores of the 
island and a few similar spots in the northern part of the lake 
rue rupicolous species. i\yQ,t large and wcll- developed shells are to be obtained (com- 
pare figs. 3A and 3B, pi. III). Such shells often reach a length of 32 mm. and a 
maximum breadth of 14 mm. Looking down through the clear water from a boat 
one sees enormous numbers of fine examples covering the rocks to a considerable 
•depth. It was only on the rocks at Chikubushima that I obtained specimens of M. 
biwae. 
No Lamellibranchiata have been found on rocks or stones in the lake. 
Melania liberiina and Bithynia striatula japonica are not in quite the same bio- 
logical category as the species discussed above. They are found, the latter in very small 
numbers only, on stones at the edge near Otsu. The speci- 
Doubtfui species. mens of M . liber Una are pale in colour and have distinct 
I The difference is so slight (less than i metre in 50) that it does not matter whether we nse English or Japanese 
feet. One Japane.se foot (shaku) equals 11 '9305 inches. 
