74 On Stocking. 



as against 220 for 22,000 yearlings the difference is 

 . perhaps hardly worth the risk but with a good man 

 30,000 yearlings might be reasonably produced, and the 

 surplus either sold to reduce the cost, or if the bottom 

 feeding turns out as I expect, then 30,000 yearlings would 

 by no means be an excessive stocking. 



I have not suggested a hatching house, because all 

 experience is against small hatcheries, they are in the end 

 more expensive than buying from larger establishments, 

 and the first structural cost is very heavy, if good in- 

 cubation is to be obtained. The principal hatching-house 

 at Howietoun cost, fitted, over 3000. Nor have I sug- 

 gested two-year-old ponds, as the depth and size, from 10 

 to 12 feet deep, and for 12,500 two-year-olds an area of 

 that depth of half an acre, makes these ponds very costly, 

 especially as if the two-year-olds are to thrive the pond 

 must be deepest and drained from the centre ; if any 

 other form of pond is used, it must either be larger in 

 extent, or much more lightly stocked. 



BREED OF TROUT. 



All the Scotch trout with two exceptions (Islay and 

 Orkney), pass from one form to another so as to render, 

 after a few generations, identification very difficult. I 

 therefore lay more stress on the selection of breeders than 

 on the original strain. But for the Lake of . . ., 

 the Loch Leven strain is, I believe, the best. Those 

 which have been bred at Howietoun since 1874 from the 

 best yearlings the produce of the largest breeders are 

 more stable in their characteristics than trout bred chiefly 

 from Loch Leven, where promiscuous breeding is tend- 

 ing to slightly lower the average weight, while a little 

 skill would in a few years very materially increase it. 



In closing this preliminary report, I have to impress 

 the importance of first killing down the pike before they 

 spawn in March. Should rearing ponds be determined 



