32 Transactions of the Canadian Institute. [Vol. IX. 



August and continues until the early part of October, although it may be 

 prolonged into November or even later when, as happens in some years, 

 the run is continued later than usual. There are some slight differences 

 in the spawning habits of the different species; the Sockeyes, for example, 

 spawning principally in lake-fed streams, while the Spring salmon and 

 Humpbacks seem to prefer streams which do not issue from lakes. But in 

 general the process is essentially the same in all the species. The female 

 chooses a situation where the current is fairly rapid and passes over a 

 gravelly bottom, forming what is locally known as a "riffel." Placing 

 herself at the upper end of the riffel, with her head pointing upstream, she 

 extrudes a number of ova and then moves away, the male immediately 

 taking her place and shedding a portion of his milt. The female again 

 returns to her original position and deposits more eggs and so the process 

 goes on day and night for one or even two weeks. In their struggles to 

 set free the reproductive elements the fish usually excavate a more or less 

 distinct hollow in the gravel of the bed of the stream, and this has fre- 

 quently been taken to be a nest prepared for the reception of the eggs. 

 As a matter of fact the current usually carries the eggs below the so-called 

 nest, and it seems more in accordance with the facts to regard the "nest" 

 as a mere accidental hollow produced by the exertions of spawning.* 



The ova, when extruded, are carried a short distance downstream, 

 but they quickly settle to the bottom, lodging between the particles of 

 gravel, and if fertilized, they develop in that situation. They are rela- 

 tively large ova, about quarter of an inch in diameter, and are of a bright 

 orange colour and hence are rather conspicuous, so that they fall an easy 

 prey to other fish, such as trout, and to water birds. Under ordinary 

 circumstances the young fish from the eggs deposited by the earlier running 

 salmon hatch out in about two and a half or three months, but those 

 deposited by the later fish may require four or even four and a half months 

 before they leave the egg, the difference being due to the lower temperature 

 of the water. When the young fish is first hatched the yolk, which forms 

 the main bulk of the egg, is not all absorbed, but distends the intestine 

 and ventral body wall into a globular projection and greatly interferes 

 with the swimming powers of the alevin, as the fish at this stage is termed. 

 The bright orange colour of this unabsorbed yolk also renders the alevin 

 a very conspicuous object when it emerges from under cover of pro- 

 tecting pebbles, and as the young fish on occasion gathers in schools they 



*In this description of the spawning process I have followed closely the account 

 given by Rutter (/. c.) for the Spring salmon of the Sacramento River, but from what 

 is known regarding the Sockeye it seems that the process is essentially the same in that 

 species. 



