274 MR- A. H. REID 



section should be attracted to or destroyed by the neglect 

 or intent of an adjoining neighbour. It is well known 

 that those portions of a river where trout spawn are not 

 favoured by them as a habitat during the rest of the 

 year, and this is probably one of nature's means of pro- 

 tecting the new fry from the depredations of the mature 

 fish. Trout would probably be hatched in one section, 

 developed in another, to spend their years of maturity in 

 a third, migrating each year to spawn in the first. There 

 are few rivers that afford spawning, rearing and feeding 

 grounds within reasonable distance of each other, and 

 it is well so, as such rivers provide but a poor 

 prospect for the small fish. Generally speaking 

 the spawning beds are situated far up in the 

 high waters or in tributary streams which afford 

 suitable gravelly beds and are too small for the sustenance 

 or comfort of large fish. Up to the present our South 

 African head waters have the advantage of being un- 

 polluted, having clean gravel beds, which is most 

 important, as mud in the gravel prevents the free per- 

 colation of the water, fouls the outer shell of the eggs 

 and may produce gases that kill an embryo so dependent 

 on a plentiful supply of oxygen in it's development. As 

 remarked elsewhere, the embryo remains in the gravel 

 from 3 to 6 weeks, according to the temperature and the 

 condition of the water, so it Avill be readily understood, 

 how desirable it is to have an unrestricted flow of clean 

 ^ater through clean gravel. A healthy, well developed 

 trout should shed about 800 eggs to the pound weight- 

 Careful observation of the " redds '' proves that a very 

 large proportion of the naturally deposited eggs are 

 injured by compression or abrasion by the gravel, 

 especially if it be sharp. 



Pollution Mortality. 



The matter of the pollution of streams in the vicinity 

 of towns, villages or factories is fast becoming one of 



