70 
The two types are distributed among the different islands 
as follows: 
Islands. 23 Rows. 21 Rows. 
FIG MENel Soocoqdupod oc Coos doocndonadu cog condos eAetnndoASshors 0 6 
IRUNGG): segestodods Boos 2 22 
Hachijoshima 0 0 
Awa lifer ocr fies 0 1 
Platdiee)iqh! SRA caso AomdspounoUL soup adabduoneDcoubdcevoSuncabooscoes 1 1 
[SGhiSate aa paqnaayopaacsoduodcodss Doo ude Sh odoootacuetoommacuadgaG 0 1 
VERE bey (ID Peosoddno osoconponsoGdtoosocdonadaoos ponsea4p004 1 0 
Wiehe abest! (UUM SossoneadapodoscabnoupoasHadgbobuooudasupadoaa 1 0 
BUSTS Dad eae pave re pete teverseksteriesteatstaverstaresetsts|cfoverelotelelsrelatatoscte totale sVatiatedatet-fateg 3 2 
Hv wbetek | SASasaoesoosoEpaU dD odDoUdoNd dnoodnoneonsbaecoDOdmeNeS 0 3 
TTNoot et imtavaleteyenetebevel stsiege es. <Velisks 'oy-7anefayetsFs akelcielabelecevel aver ateveralolstetohal= leis terenet ute 8 38 
VARIATION IN PHYSIOGNOMY. 
There are marked differences in the profile and contour 
of the head of certain specimens. The snout may be short and 
decidedly tilted at the tip. This is very striking but is not 
associated with any other constant character. Specimens with 
the turned-up snout have even been captured in the same 
thicket with the commoner variety having the flat head. The 
other character, the contour of the muzzle, varies in being 
more acute in some specimens than in others. With just a 
few specimens at hand it would not be unreasonable to assert 
that two distinct species were under consideration. But, after 
examining a large series and finding every combination of 
these characters one is convinced of the specific identify of the 
snakes. From this it appears that not only may a single spe- 
cies of serpent vary in details of squamation, but also struc- 
turally to the extent of producing a marked difference in its 
external anatomy. 
Structural variation in the size of the eye is known to 
exist among specimens of Thamnophis parietalis (Say) 1823, 
taken in California. The extremes are astonishingly wide 
apart, yet a perfect series of intermediate diameters may be 
found. 
VARIATION IN PREFRONTALS. 
There is considerable variation in the shape of these 
shields. They vary from being as long as wide to a shorter 
condition in which the length is but 2-3 of the width. The 
specimens with the shorter shields have the shorter and 
broader snout. Two percent of the examples studied showed 
the tendency to fragmentation of this shield. 
VARIATION IN FRONTAL. 
In the majority of specimens from Shanghai the frontal 
shield is about 1 2-5 times longer than broad. In all the 
(1) 1903, Wall, Proc. Zool. Soc. London, p. 102. 
