242 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1946 
New Zealand in autumn, and return southward toward the Antarctic 
in spring, and Harmer (1931) and Matthews (1937) show that the 
monthly catches off Angola and Durban reach a peak first in July 
and then again in September or October, whereas at the French Congo 
there is a single peak about July and August. These facts accord 
with the statement that the humpbacks mostly pass the subtropical 
coasts in the earlier part of the season, and that after some of them 
have penetrated as far north as the Equator, they return past the sub- 
tropical coasts toward the end of the season. From the authors 
mentioned above, and from Kellogg (1929) and the International Sta- 
tistics, we find that humpbacks are hunted mainly in tropical latitudes 
on both sides of each of the southern continents in winter, and in the 
open ocean in the Antarctic in summer. Conclusive evidence that a 
long-range migration takes place is provided by whale marking, for 
a number of marks fired into humpbacks off the pack ice in high lati- 
tudes have been recovered by whalers off the northwest coast of Aus- 
tralia, and some also off Madagascar (see Rayner, 1940). It is prob- 
able that the majority of humpbacks undertake this migration annu- 
ally, but it does not follow that all of them penetrate far into the 
Tropics, and indeed some appear to remain in the Antarctic, for some 
rare examples have been taken in Antarctic waters at a time when 
whaling was continued on a small scale through the winter at South 
Georgia (Risting, 1928). 
The marking of whalebone whales in the south has shown that after 
they have migrated northward they usually return to the same part 
of the Antarctic in the following summer. This applies especially to 
humpbacks. Hjort, Lie, and Ruud have published data on the re- 
gional distribution of the Antarctic catches of the pelagic whaling 
fleet. An analysis of their figures (1938, 1939), together with the 
marking records of the Discovery Committee, shows (Mackintosh, 
1942) that in summer in the Antarctic the humpbacks are segregated 
into clearly separate groups in positions which seem to correspond 
with the separate tropical coastal resorts in winter. Recoveries of 
whale marks demonstrate that an Antarctic group lying southwest 
of Australia in summer contains the same whales as appear off the 
west Australian coast in winter, and a connection has been estab- 
lished between a group south of South Africa and the humpbacks 
caught in winter off Madagascar. It thus seems highly probable that 
each of the Antarctic summer groups has its own migration route to 
the coastal waters of a continent lying approximately to the north of 
it (see fig. 1). It is to be supposed therefore that the southern stocks 
of humpbacks are divided into several communities which are for the 
