190 
Part III. — Tenth Annual Report 
relation to inland fisheries, has reached the perfection of art at the cele- 
brated establishment at Howietoun. But pisciculture as applied to sea- 
fish is of quite recent date. Artificial fecundation and hatching of the 
eggs of certain sea-fishes has been accomplished on an experimental scale by 
many naturalists — by Sars in Norway, Professor M'Intosh and others in 
this country, by the German scientists on the Baltic, and in America. 
But the first to work at the subject successfully, from the practical point 
of view, appears to have been Captain G. M. Dannevig, the Director 
of the well-known sea-fish hatchery near Arendal, in Norway, and 
who is the greatest authority on the subject. His institution was 
established seven or eight years ago, and has since been enlarged. More 
recently, a hatchery for sea-fish was established in Newfoundland, and 
another in Canada ; and at Wood's] Holl and other stations in the 
United States the Fish Commission have engaged in sea-fish culture for a 
considerable number of years. 
Before describing what is being done in Scotland, a word or two may 
be devoted to the enunciation of the principles upon which sea-fish 
culture rests. Owing to the intensity of the struggle for existence in the 
sea, the destruction which takes place at various stages in the life of 
fishes is immense. Hence, most sea-fishes produce an enormous number 
of eggs.* A single female turbot may produce in one season nine or ten 
millions ; a cod, six or seven millions ; a ling, twenty or thirty millions ; 
a haddock, half a million to about a million, and so on. The import of 
this enormous fecundity has frequently been altogether misunderstood ; 
arguments have been based upon it to show the inutility of interference 
in fisheries. In reality fecundity is a measure of the natural 
destruction that occurs in the life-history of any species, since, on the 
reasonable assumption that the total number of a species remains fairly 
constant over a period, it is only necessary that a few individuals of the 
new generation should, on the average, survive to the reproductive stage 
in order to keep up the relative abundance of that species. Hence the 
proportion of the eggs produced by sea-fishes which give rise to reproduc- 
tive individuals is infinitesimal. Of the ten millions produced by the 
turbot, 9,999,998, or thereabout, take no part in the production of another 
generation, but are destroyed at one period or another when left to 
natural conditions. So with other species. Now, since the fundamental 
principle of pisciculture is the protection of the organism at its earlier 
stages from destruction by its natural foes, it would be desirable, from 
the most practical point of view, to know at what stage or stages this 
enormous destruction mostly occurs. Is it chiefly in the egg state, or in 
the larval, post-larval, or ' immature ' periods 1 There are some facts of 
importance in considering this point. In the first place, as a general rule, 
the fecundity of species whose eggs are buoyant or pelagic, is consider- 
ably greater than the fecundity of species whose eggs are demersal or 
attached ; in other words, the ratio of the destruction is much greater 
among the former. The number of ova produced no doubt depends to some 
extent on the size of the essential parts of the ovum and the rate of growth 
of the species ; but, taking these factors as far as possible into account, 
it appears that the fact is as stated. The herring, for instance, produces 
about 30,000 eggs; the pogge (Agonus), about 1200; the lumpsucker, 
over 100,000. Some herring, possibly, reach sexual maturity within 
a year, but probably later in the great majority of cases ; the lumpsucker 
probably takes at least as long as the cod. The destruction, therefore, 
* Vide my paper on 'The Comparative Fecundity of Sea-Fishes/ Ninth Annual 
JZeport. 
