1248 METABOLISM OF LIPOIDS, STEROLS, [pt. iii 



of sea-water to be, for purposes of calculation, 0-034 mgm, per litre. 

 Now during development we find in round numbers an increase of 

 from 7-5 to 12-5, i.e. 5-0 mgm. per gm. dry weight of eggs, and this, 

 reckoning about 4,000,000 eggs to the gram, would be 0-012 x io~* 

 gm. Comparing this estimate with that for the amount of phosphorus 

 in the sea-water, it will be evident that the requirements of the 

 embryo will be amply met by the water with which it comes in con- 

 tact. In addition to this, turbulence effects must be remembered, and 

 it is probable that under favourable conditions the embryo could 

 dispense with a great deal of its activity as far as the phosphorus- 

 intake is concerned. But if it floated motionless in a perfectly still 

 medium containing phosphorus in as high concentration as sea-water, 

 it is likely that the supply would be deficient. 



These facts have a bearing on the problem of the origin of fresh- 

 water fauna. The classical theory of Sollas attributes the paucity of 

 fresh-water species mainly to the custom, usual among marine in- 

 vertebrates, of hatching early in a larval form and dispersing in the 

 plankton. These minute ciliated larvae, he pointed out, cannot swim 

 against currents of any magnitude and are ill-adapted for river life, 

 so that even if a marine animal solved the osmotic difficulties asso- 

 ciated with existence in fresh-water it would see its eggs and larvae 

 swept out to sea again as fast as it produced them. The well-known 

 phenomena of poecilogony (see p. 316) favour this view. But such a 

 theory has no explanation to give for the cases (e.g. the cephalopods) 

 where the embryo spends a long time in its egg before hatching out 

 substantially mature, and yet the group has never penetrated into 

 fresh- water (see p. 317). Now if eggs such as these depend on the 

 salts of the sea-water as a kind of additional yolk, the inorganic en- 

 vironment may become an important limiting factor, and in the case 

 of phosphorus in particular, the margin between supply and demand 

 in normal sea-water may be much reduced. Thus the plankton may 

 use up nearly all the available phosphorus, for as Atkins has shown, 

 there is a seasonal variation in the English Channel, surface water 

 containing 0-0162 mgm. P per litre as a winter maximum and only 

 0-0032 as the summer minimum, the difference being associated with 

 the growth of the phytoplankton. Corresponding figures for the 

 Clyde area, given by Marshall & Orr, are 0-032 and 0-0022 mgm. P 

 per litre respectively. In the Pacific Ocean the same cycle goes on, 

 according to Stanford and Moberg. Finally, the analyses of Juday, 



