102 ANIMAL LIFE 
but a year, ascending the rivers in numbers when young in 
the spring, the whole mass of individuals dying in the fall 
after spawning. 
Naturalists have sought to discover the reason for these 
extraordinary differences in the duration of life of different 
animals, and while it can not be said that the reason or 
reasons are wholly known, yet the probability is strong that 
the duration of life is closely connected with, or dependent 
upon, the conditions attending the production of offspring. 
It is not sufficient, as we have learned from our study of 
the multiplication of animals (Chapter III), that an adult 
animal shall produce simply a single new individual of its 
kind, or even only a few. It must produce many, or if it 
produces comparatively few it must devote great care to 
the rearing of these few, if the perpetuation of the species: 
is to be insured. Now, almost all long-lived animals are 
species which produce but few offspring at a time, and 
reproduce only at long intervals, while most short-lived ani- 
mals produce a great many eggs, and these all at one time. 
Birds are long-lived animals; as we know, most of them 
lay eggs but once a year, and lay only a few eggs each time. 
Many of the sea birds which swarm in countless numbers 
on the rocky ocean islets and great sea cliffs lay only a 
single egg once each year. And these birds, the guillemots 
and murres and auks, are especially long-lived. Insects, on 
the contrary, usually produce many eggs, and all of them 
inashort time. The May-fly, with its one evening’s lifetime, 
lets fall from its body two packets of eggs and then dies. 
Thus the shortening of the period of reproduction with the 
production of a great many offspring seem to be always 
associated with a short adult lifetime; while a long period 
of reproduction with the production of few offspring at a 
time and care of the offspring are associated with a long 
adult lifetime. 
There seems also to be some relation between the size 
of animals and the length of life. As a general rule, 
