THE ANIMALS OF LAKES AND RIVERS 187 
growth of the more delicate animals, and must be 
a great bar to the passage of marine animals up estu- 
aries and rivers. Almost any coast-line will illustrate 
the same thing on a smaller scale. The shore collector 
knows that to get the rarer and more delicate animals 
he must seek rocks or beaches remote from the mouths 
of streams, for at the immediate outlets of these only 
the hardiest forms occur. This statement is not in- 
consistent with that already made, that the waste of 
the land is an important part of the food of the sea- 
animals, for oxygen is a prime necessity which precedes 
even the need of food, and few animals can take in 
oxygen if their respiratory organs are choked with mud. 
But it is not only the suspended matter in the water 
of lakes and rivers which is inimical to animal life. 
The presence of large amounts of lime or manganese 
salts, or of humic acid, may render the water unsuitable 
to certain forms. 
Still another point of great importance is the enor- 
mous variation in temperature to which lakes and 
rivers are subject. These variations are normally so 
great that they render the water absolutely unsuited 
to stenothermal animals. All fresh-water forms must 
be eurythermal (cf. p. 160), and temperate and polar 
forms must have special means of protecting them- 
selves, or their offspring, against the cold of winter. 
In the ‘ lakes’ of Victoria Land, which are almost per- 
manently frozen, the naturalists of Sir Ernest Shackle- 
ton’s expedition found rotifers which can apparently 
tolerate being frozen into the ice for years, and will 
yet hatch out if the temperature rises and the ice melts. 
This is an extreme case, but it suggests the adaptations 
necessary before animals can thrive in small masses of 
water. Low temperatures are not, however, the only 
