544 CONCLUSIONS. 
4°. We observe that in one month not all the appearances took 
place in the same latitude, consequently in one and the same month 
they are scattered over a vast portion of the ocean. 
5°. When the migration from north to south begins, which of 
course must be influenced by the early or late setting in of autumn, 
it seems that not only the mdividuals which have proceeded to the 
most northern coasts of Norway, but also some other individuals 
begin their migration towards the south. I think that we must find 
in this fact the explanation that even in July appearances took place 
at from 5° to 38° southern latitude, and that on August 6 an . 
individual was seen at lat. 24° S. swimming ¢owards the S. W. 
Though I have no appearances in the South Atlantic in the 
month of October, | am convinced that the greater part of the in- 
dividuals are there during this month, as well as in November, 
December and January. — The reason that there are so few reports 
from these regions is of course that in comparison with the North 
Atlantic, a far smaller number of vessels visits the South Atlantic. 
The two appearances which happened in the Pacific, and of which 
the dates are mentioned, are: 
April. 
South of Japan 81° N. (151) 
May. 
South of Australia 438°S. (122) 
And those of the Indian Ocean: 
January. 
Gulf of Aden 12° N. (149) 
March. 
Geographe Bay 333)" Sh (150) 
May. 
Indian Ocean Q°N., (147) 
September. 
Straits of Malacca 3° N. (146) 
Indian Ocean 157 Se (1238) 
Suppose that some individuals in the Atlantic migrate towards 
the south beyond the latitude of the Cape of Good Hope and get 
much farther than 20° eastern latitude, they will come into the Indian 
Ocean. I think that when these individuals returning to the north, 
find themselves checked by the continent of Asia, they will swim 
in any direction, and that perhaps most of them will find back 
the outlet round the Cape of Good Hope or south of Australia, so 
that in such cases individuals will be met with in the South At- 
