1909 THE ANIMALS OF LAKES AND RIVERS 
swimming larvae called planulae; Hydromedusae 
produce free-swimming medusoids ; polychaete worms 
have free-swimming larvae called trochospheres, all 
delicate forms incapable of swimming against currents. 
These particular ponds have at intervals a connexion 
with the sea, and so can be peopled afresh from it, but 
if we suppose that at some future period the connexion 
with the sea were to be reduced to a swift-flowing 
stream, could we suppose that the animals would 
persist? Obviously not, for the medusoids, the 
planulae and the trochospheres would tend to be swept 
out to sea by the stream, and there would be thus no 
young forms to replace the parents when death took 
place. So real is this danger that, with rare excep- 
tions, fresh-water organisms have no free-swimming 
stage in their life-history. Only those marine animals 
could colonize fresh water which had no plankton 
stage in their life-history, or which managed somehow 
to survive until this free-living stage could be sup- 
pressed by progressive variation. This perhaps gives 
us a second reason why so many animals with a 
terrestrial ancestry have succeeded in colonizing fresh 
water, for these forms would not have a free-swimming 
stage. Thus while most marine gasteropods have free- 
swimming larvae, fresh-water snails, like land snails, 
hatch as miniature adults, which makes life in ponds 
and streams easy for them. These arguments of course 
only apply to those delicate organisms which are in- 
capable of swimming against currents. Relatively 
powerful swimmers, like many insect larvae, can 
resist moderate currents, though they avoid regions 
where the water is in rapid movement. 
But, it may be said, in a large lake the need for dis- 
tribution is as great as in the sea, how do the lake 
