16 Mr, C. T. Regan on the British Fishes 
“During the colder months of the year they are frequently 
found in the stomachs of large ground-fish that have been 
taken with ground-lines some distance from the shore” ; 
whilst Kishinouye writes of S. sagax, ‘‘ We are told by 
fishermen that the sardine is often found in the stomach of 
deep-sea fishes, such as mutsu (Scombrops chilodipteroides), 
gisu (Pterothrissus gisu), and tara (Gadus brandti).” 
It does not seem likely that under present conditions there 
is any interchange between the different colonies of Sardina 
sagax, Chilean, Californian, Japanese, and South African ; 
but under somewhat different conditions this may not have 
been the case—with a warmer northern climate Sardina sagaz 
may have had a continucus range in the northern part of the 
Pacific. On the other hand, a contraction of the Tropical Zone 
on the Pacific coast of America may have permitted an inter- 
change of Chilean and Californian pilchards migrating in 
moderately deep water. SS. neopilchardus may have evolved 
from S. sagax in the seas of Australia and New Zealand, 
whilst the latter has persisted without such conspicuous 
modifications in the other parts of its range. 
Alosa has a much more restricted distribution than Sardina, 
as it occurs only in the North Atlantic and the Mediterranean. 
Southwards it extends a short distance into the Tropical Zone 
and northwards it ranges in the summer months well into. the 
Subarctic Zone. 
The more restricted distribution and the greater tendency 
to form local races of Alosa as compared with Sardina are no 
doubt correlated with the biological difference between the 
two genera, that the shads breed in fresh water and the young 
live in the rivers for a year or two before migrating to the 
sea, whilst the pilchards are strictly marine and have pelagic 
eggs and young. The local forms with a most restricted 
distribution are, of course, the non-migratory lacustrine 
colonies, such as the shads of Killarney or of Lake Garda. 
Clupea harengus of the North Atlantic is closely related to 
C. pallasii of the North Pacific. Both species extend into 
the Arctic Zone, but neither is found on the northern coasts 
of Asia or America; southwards they range throughout the 
Subarctic Zone, and somewhat overstep its southern boundary. 
There can be little doubt that under somewhat milder climatic 
conditions the ancestral herring ranged along the whole 
northern coast of Eurasia and, perhaps, of America ; it is not 
surprising that the herrings, belonging to the colder northern 
seas, are not represented in the Southern Hemisphere. 
Clupea sprattus, of Western and Southern Hurope, has a 
more southerly distribution than C. harengus, but does not 
