# of the Subfamily Clupeinz. 15 
TI. GEOGRAPHICAL DISTRIBUTION. 
In the report on the ‘Terra Nova’ fishes I gave a general 
account of the distribution of Antarctic and Subantarctic 
fishes, and I came to the conclusion that to illustrate the 
distribution of coast-fishes south of the tropics three zones 
might be recognized—Antarctic, Subantarctic, and South 
Temperate; the northern boundaries of these are the mean 
annual surface-isotherms of 6°, 12°, and 20° C. respectively. 
In the Northern Hemisphere the problem is more compli- 
cated, and it is more difficult to limit and define the zones of 
distribution ; but, if we regard the isotherms of 12° and 20° C. 
as bounding the Temperate Zones both in the north and the 
south, we find that the genus Sardina may be described as 
inhabiting the North and South Temperate Zones, barely 
overstepping their limits (cf. map, Pl. II.). 
Sardina is absent from the Western Atlantic and from 
China and the west of Corea; the reason for this is unknown, 
but in the Western North Atiantic the meeting of the 
Labrador Current and the Gulf Stream produce a sudden 
transition from subarctic to almost tropical conditions, and 
the absence of Sardina may be connected with this. 
The wide distribution of this genus is of interest when the 
close relationship of the species, and especially the practical 
identity of the pilchards of South Africa, Japan, California, 
and Chile, is taken into consideration. 
The species are not oceanic, but may be found 50 miles or 
more from land. ‘The eggs are pelagic and the young fish 
swim at the surface, so that one can understand how a South 
American species may have reached New Zealand or South 
Africa ; but the crossing of the Tropical Zone would be more 
difficult. 
There is good evidence that in comparatively recent times 
the limits of the Temperate Zones have fluctuated considerably. 
‘Lo take only one example, the trout (Salmo trutta) of the 
Atlantic coast of Europe ranges southward as a marine fish 
to the Bay of Biscay, not much. south of the normal northern 
limit of the pilehard (Surdina pilchardus). But the presence 
of a trout in the rivers of Morocco indicates that not long ago 
sea-trout ranged as far to the south as the pilchard does now, 
and at that time the ancestral pilchard may, perhaps, have 
extended into the area of the present Tropical Zone. 
Fishes that descend to considerable depths are less likely 
than shallow-water species or surface-swimmers to find the 
‘Tropical Zone an impassable barrier. It is therefore of some 
interest to note that Day has written of Sardina pilchardus :— 
