F. DAY: PERIODS OF MIGRATION. 



121 



fish,* with ovaries 47 inches in length, the two weighing 2^ ounces, 

 and each egg being o*i inch in diameter. It is clear that a salmon 

 having eggs one-tenth of an inch in diameter at the end of November, 

 could not have its ova sufficiently ripe to spawn within the next two 

 months, while experience tells us that no other period for depositing 

 eggs will normally come round before this time the succeeding year. 

 It has long been accepted as a physiological necessity that a female 

 smolt must descend to the sea before it can develop eggs, the reason 

 advanced being that the development of ova requires far more 

 nourishment than that of milt ; that in the ovary of the female the 

 eggs are formed nearly simultaneously, and their development is 

 uniform, one being enveloped in as large an amount of albumen as 

 another. But in order to produce this albumen, a far greater quantity 

 of food is needed than the fish can normally procure in freshwater 

 rivers. On the other hand, grilse at Howietoun, both last year and 

 during the present season, have given eggs without going to the sea ; 

 and also the land-locked salmon breeds in fresh water without 

 descending to the ocean. 



I think that the explanation of these apparently contradictory facts 

 is possible. Thus, it is generally admitted that salmon, while residing 

 in rivers, do not increase in weight, but rather fall off the longer they 

 are absent from the sea, existing as they mainly do upon the fat which 

 they have accumulated while feeding in the salt water, and such food 

 as they can procure sufficient for nutrition of the body, but insufficient 

 in females for breeding purposes. If this is so (and of it I think there 

 can be no reasonable doubt), they would be unable to obtain enough 

 nourishment wherewith to develop eggs so long as they continued in 

 the river, that, in short, they could not do so without another visit to 

 the sea ; consequently these early-ascending salmon, until they have 

 again descended to the salt water, cannot be those fish from which 

 we have to expect ova for replenishing the stock in our rivers. Knox, 

 in 1854, observed that a smolt, after first descending to the ocean 

 and tasting its marine food, never again resorts to its infantile food as 

 a constant mode of nourishment. 



This brings me to the question of how it is possible to prove that 

 insufficient nourishment can impede or prevent spawning among the 



* Three similar ones were likewise captured that clay, but being uninjured were 

 at once returned to the Teith. This must not be considered an exceptional 

 occurrence, as clean fish at this time are invariably netted when seeking for gravid 

 salmon. In the Rhine, Barfurth observed in 1874, that spawners ascend from 

 September to November, while there is likewise a barren winter variety coming 

 sporadically and for a brief season from September until Ma)'. 

 April 1886. 



