512 TRANSACTIONS OF SECTION D. 
Having now stated the facts, we proceed to try an explanation of these, 
confining ourselves to the Atlantic species. It has been shown that the 
Atlantic fresh-water eels live in tropic, in warm, and in cold temperate, 
even in Arctic regions; but just on account of this astonishing power to 
submit to the most varied outer conditions their distribution appears some- 
what incomprehensible. Why do they stop suddenly southwards so that 
the greater part of South America and West Africa axe entirely destitute 
of eels in spite of the fact that some of the largest and most fish-abundant 
fresh-water systems are found here, and many places apparently would 
offer excellent conditions for their thriving? And why do they occur in the 
oceanic islands, in which fresh-water fishes otherwise are lacking ? 
Before answering these questions, and some others to be mentioned later 
on, it will be necessary to refer to some of the recent results of sea-investi- 
gations. It is now a well-ascertained fact that the sensibility of a species 
of fish to its surroundings may vary a great deal during its growth and 
at its spawning-time; during the latter its requirements as to depth, 
temperature, and salinity may be very different from those during its 
growth; therefore its distribution during spawning may often be very 
different from that during the growth; and it may be the conditions for 
spawning which are the chief factors in determining the distribution of 
the species afterwards. Investigations made since 1904 with the Danish 
investigation steamer Thor and the Irish Helga in the Atlantic west of 
Europe have shown that the European fresh-water eel propagates in the 
Atlantic far from the coasts, and that its propagation requires large depths 
(at least 1,000 metres), besides a high salinity, more than 355'20 per cent., 
and a high temperature of the water, at least 7° C., at a depth of 1,000 
metres. These facts render the distribution and migration of the eel in 
the northern part of Europe quite comprehensible. 
To apply these conditions to the whole region. On looking at a chart 
on which the temperatures at a depth of 1,000 metres are given, the distri- 
bution of the Atlantic fresh-water eels suddenly becomes quite clear. We 
see that the 6° and 5° isotherms run from the West Indies eastwards over 
to Cape Verde, so that the whole southern part of the Atlantic has a 
lower temperature at 1,000 metres depth. Off the greater part of Brazil 
the temperature is even only between 3° and 4° C., off the west coast of 
Africa a little higher, for the greater part between 4° and 5°. On the Pacific 
coast of America the temperature in deep water is also low, in no place 
reaching 7°, being in the greater part below 5°. We understand at cnce 
that the want of eels in all the large fresh-water systems of South America, 
the western North America, and West Africa is due to the fact that the 
temperature of the adjacent seas is too low to permit of the propagation. 
Therefore no immigration of young eels to these regions is possible. We 
may say, although it sounds almost paradoxical, that it is the cold which 
prevents these regions, which contain some of the warmest countries upon 
earth, from being inhabited by fresh-water eels. 
On the East of Africa, in the Indian Ocean, as well as on the South of 
Asia, we meet with a higher temperature. And here again fresh-water eels 
are found, though not the same species as in the Atlantic region. 
Taking for granted that the eels propagate out in the sea, we under- 
stand why they may be present in oceanic islands, where no other fresh- 
water fishes are living, and looking at the temperature curves we under- 
stand why only the oceanic islands lying north of the Cape Verde Islands 
are inhabited by eels; only these lie inside the region where the temperature 
in the depth is high. 
1 Here it may be remarked that high temperature and high salinity are always 
connected in the great depths of the oceans. Thus it is impossible to ascertain 
whether the one or the other factor, or koth combined, are determining the propaga- 
tion of the eel, i 
