126 SEASON 1874-75. 



box, the tubes were cut to 8 inches, and placed across the box, as 

 now used, but the fastening was even more troublesome than the 

 last. I spent many hours with a small thumb -drill and a large 

 school slate. I cut a narrow slip off the slate the length I required 

 for the inside of the grille frame, and with a pair of compasses I 

 divided it accurately ; and with the thumb-drill I bored holes 

 opposite the end of each tube, and when a pair of these slips were 

 ready, I sewed with fine wire the tubes to the slate. A quarter of 

 the sides of the frame of the grille were cut away, forming a 

 ledge for the slate. Thus the wood covered the wire sewing, and 

 there was no necessity to use silver wire ; brass wire would kill 

 any ovum in contact with it. Stone says, in the Domesticated 

 Trout, p. 55, " Fourteen trout eggs were placed on a copper 

 screen, in November 1869, at Cold Spring Trout Ponds, and in fifty 

 days they had absorbed so much copper that they were of a dark 

 brown tinge, and hard as peas." Grilles made in this manner were 

 an impossibility for work on more than a merely experimental 

 scale, but had so many advantages that I persevered ; and the next 

 improvement was the substitution of perforated zinc for the slate. 

 Two strips were cut off a sheet through contiguous lines of holes, 

 and one was fixed in the check of the frame, and the other into a 

 fillet. The fixing was simple, merely a draught with a sharp brad- 

 awl, and the zinc pressed in. It held as long as the wood lasted 

 sound. But the labour of arranging the glass tubes was enormous. 

 and if one at that time a by no means infrequent chance got 

 broken, the whole had to be taken out before it could be replaced. 

 One day I was carrying over a broken grille to the carpenter's shop 

 (a distance of nearly two miles from Middlethird), when the idea 

 struck me, Why should not the holes in the zinc remain whole ? 

 The apparent answer was, that any size of perforated zinc large 

 enough to admit the glass would keep the tubes so far apart that 

 the eggs could drop through. This was so, but I made one to 

 try. The small ova of the burn-trout passed through with ease, 

 not to say rapidity ; the eggs of the Loch Levens behaved better ; 

 some, not many, remained on the grille. The problem was solved 

 (Fig. 68). I had only to use a smaller size zinc, and widen the 



