86 DOMESTICATED TROUT. 



5. Absence of fixed hiding-places. Such hiding-places 

 are bad, because in general you do not know what the 

 fish are doing in them. They may be dying there. 

 They may be crowding in them too thickly. The 

 hiding-places may conceal a snake, or a frog, or a 

 cannibal trout, which is making deadly havoc 'with 

 the small fish. The best rule for hiding-places 

 is, not to have them fixed, but so that you can 

 always remove the shelter, and look in if you 

 wish ; then you will know what is going on in 

 them. 



6. Compactness of the fish themselves for feeding. This 

 I regard as a very important point. When the very 

 young fry are scattered widely apart, you are obliged 

 to feed them at a great disadvantage. 



Only a few can be got together in a spot to feed. 

 Sometimes they are so shy that they can only be fed 

 at all with great difficulty, and, with the best you can 

 do, nineteen twentieths of the food will go to the bot- 

 tom. On the contrary, when the fish are compactly 

 confined, their numbers seem to give them confidence, 

 and they do not attempt to run away from the food. 

 They will gather together to get it, instead of scatter- 

 ing as before ; and, being so thick together, they will 

 consume nearly all the food given them, and very little 

 will go to the bottom. The advantage gained by this 

 is very great. 



7. Protection against outside enemies. The necessity 

 of this protection is obvious. It is obtained by at- 

 taching to the rearing box a wire-work lid, fitting 

 down tightly, and provided with a padlock. 



