4 HISTORY OF HOWIETOUN. 



the fact that the spawning beds of the herring are a resort of 

 many and ravenous fishes, I do not feel inclined to attribute any 

 serious proportion of loss. The third period is that immediately 

 succeeding the absorption of the umbilical vesicle. The ova of 

 Salmones, and most of the ova of Salvelini hatching in running 

 water, the young fry are nursed by the current which bears particles 

 forming their food more or less constantly past their resting-places. 

 and as they are hatched out generally in the natural streams 

 so late in the spring, that by the time the yolk sac is absorbed 

 there is an abundance of animalculse in the water, the loss at this 

 period is probably less than at either of the preceding ; but with 

 herring, in a great number of instances, it is easily conceivable 

 that it is precisely at this period myriads of the young fish perish. 

 The yolk vesicle probably contains little more nourishment than 

 suffices to supply the young fish on their journey from the bottom 

 to the surface of the water, and their future existence depends on 

 the quantity of microscopic larval life they find on their arrival 

 at the surface. The conditions which determine the quantity of 

 these forms are as yet very imperfectly understood, but it is pro- 

 bable that atmospheric conditions, temperature, and, above all, 

 storms, must be important factors. 



Should further research bear out these probabilities, the culture 

 of such fish as herring will differ from the culture of large-ovaed 

 Salmonidce in this, that the quantities of the latter will be 

 increased by piscifacture, 1 and of the former by legislative enact- 

 ments framed with a view to secure natural reproduction, when 

 and where the conditions (now being investigated by the Fishery 

 Board for Scotland) most favourable to the survival of the myriad 

 brood exist. Perhaps I may make the point plainer if I assume 

 a purely hypothetical case. 



Suppose the various shoals of herring to be more or less local 

 races imperfectly localised, but still sufficiently periodic in their 

 migrations to enable these to be mapped with tolerable accuracy ; 



Suppose, secondly, investigation shows that the resulting 

 brood from spawn deposited in certain places, and at certain 



1 Bulletin de MinisU-rc d' Agriculture. 



