Duration of life 257 
All these facts prove that the velocity of development of animal 
life in Arctic regions, where the temperature is near the freezing 
point of water, must be from two to three times smaller than in 
regions where the temperature of the ocean is about 10° C. and from 
four to nine times smaller than in seas the temperature of which 
is about 20°C. It is, therefore, exactly the reverse of what we 
should expect when authors state that the density of organisms at or 
near the surface of the ocean in polar regions is greater than in more 
temperate regions. 
The writer believes that this paradox finds its explanation in 
experiments which he has recently made on the influence of tempera- 
ture on the duration of life of cold-blooded marine animals. The 
experiments were made on the fertilised and unfertilised eggs of the 
sea-urchin, and yielded the result that for the lowering of tempera- 
ture by 1°C. the duration of life was about doubled. Lowering the 
temperature by 10 degrees therefore prolongs the life of the organism 
2 i.e. over a thousand times, and a lowering by 20 degrees pro- 
longs it about one million times. Since this prolongation of life 
is far in excess of the retardation of development through a lowering 
of temperature, it is obvious that, in spite of the retardation of 
development in Arctic seas, animal life must be denser there than in 
temperate or tropical seas. The excessive increase of the duration of 
life at the poles will necessitate the simultaneous existence of more 
successive generations of the same species in these regions than in 
the temperate or tropical regions. 
The writer is inclined to believe that these results have some 
bearing upon a problem which plays an important role in theories of 
evolution, namely, the cause of natural death. It has been stated 
that the processes of differentiation and development lead also to the 
natural death of the individual. If we express this in chemical 
terms it means that the chemical processes which underlie develop- 
ment also determine natural death. Physical chemistry has taught 
us to identify two chemical processes even if only certain of their 
features are known. One of these means of identification is the 
temperature coefficient. When two chemical processes are identical, 
their velocity must be reduced by the same amount if the tempera- 
ture is lowered to the same extent. The temperature coefficient for 
the duration of life of cold-blooded organisms seems, however, to 
differ enormously from the temperature coefficient for their rate of 
development. For a difference in temperature of. 10° C. the duration 
of life is altered five hundred times as much as the rate of develop- 
ment ; and, for a change of 20°C., it is altered more than a hundred 
thousand times as much. From this we may conclude that, at least 
for the sea-urchin eggs and embryo, the chemical processes which 
D. 17 
