TRAKSACTIONS OF SECTION D, 671 



Does the size depend on the number, or the number on the size ? To answer in 

 a word, I believe that the size generally depends on the number, and that the 

 number is mainly determined by the risks to which the species are exposed. At 

 least so many eggs will in general be produced as can maintain the numbers of 

 the species in spite of losses, and there is some reason to believe that in fresh waters 

 the risks are less than in the shallow seas or at the surface of the ocean. ^ In most 

 parts of the world the fresh waters are of small size, and much cut up. Every 

 river-basin forms a separate territory. Isolation, like every other kind of artificial 

 restriction, discourages competition, and impedes the spread of successful competi- 

 tors. In the shallow seas or at the surface of the ocean conquering forms have a 

 free course ; iu lakes and rivers they are soon checked by physical barriers. 



A large proportion of animals are armour-clad, and move about with some 

 difficulty when they have attained their full size. The dispersal of the species is 

 therefore in these cases efiected by small and active larvae. Marine animals (whether 

 littoral or pelagic) commonly produce vast numbers of locomotive larvae, which 

 easily travel to a distance. Floating is easy, and swimming not very difficult. A 

 very slightly built and immature larva can move about by cilia, or take advantage 

 of currents, and a numerous brood may be dispersed far and wide while they are 

 mere hollow sacs, without mouth, nerves or sense-organs. Afterwards they wiE 

 settle down, and begin to feed. In fresh waters armour is as common, for all that 

 I know, as in the sea, but locomotive larvae are rare." There is no space for eflec- 

 tive migration. Even a heavy-armoured and slow-moving crustacean or pond- 

 snail can cross a river or lake, and to save days or hours is unimportant. In rivers, 

 as Sollas has pointed out, free-swimming larvae would be subject to a special risk, 

 that of being swept out to sea. This circumstance may have been influential, but 

 the diminished motive for migration is probably more important. At least an 

 occasional transport to a new area is indispensable to most freshwater organisms, 

 and very unexpected modes of dispersal are sometimes employed, not regularly in 

 each generation, but at long intervals, as opportunity ofiers. 



Early migration by land is nearly always out of the question. Walking, and 

 still more flying, are difficult exercises, which call for muscles of complex arrange- 

 ment and a hard skeleton. A very small animal, turned out to shift for itself on 

 land, would in most cases perish without a struggle. There might be just a 

 chance for it, if it could resist superficial drying, and were small enough to be 

 blown about by the wind (Infusoria, Rotifera, and certain minute Crustacea), or 

 if it were born in a wet pasture, like some parasitic worms. 



We can define two policies between which a species can make its choice. It 

 may produce a vast number of eggs, which will then be pretty sure to be small 

 and ill-furnished with yolk. The young will hatch out early, long before their 

 development is complete, and must migrate at once in search of food. They will, 

 especially if the adult is slow-moving or sedentary, be furnished with simple and 

 temporary organs of locomotion, and will generally be utterly unlike the parent. 

 The majority will perish early, but one here and there will survive to carry on 

 the race. 



Or the parent may produce a few eggs at a time, stock them well with yolk, 



' Indications are given by the survival in fresh waters of declining groups, e.g.. 

 Ganoid Fishes, which, when dominant, maintained themselves in the sea ; and by the 

 not uncommon case of marine animals which enter rivers to spawn. I do not at- 

 tempt to count among these indications the supposed geological antiquity of fluvia- 

 tile as compared with marine animals. Some marine genera are extremely ancient 

 (Lingula, Nucula, Trigonia, Nautilus) ; a perfectly fair comparison is almost impos- 

 sible ; and great persistence does not necessarily imply freedom from risks. In the 

 Mollusca, which afford a good opportunity of testing the effect of habitat upon the 

 number of the eggs, marine species seem to produce more eggs as a rule than fluvia- 

 tile, and these many more than terrestrial species. 



- Dreyssensia and Cordylophora are examples of animals which seem to have 

 quite recently become adapted to fresh-water life, and have not yet lost their loco- 

 motive larvae. Many instances could be quoted of marine forms which have become 

 fluviatile. The converse is, I believe, comparatively rare. 



