758 THE POPULAR SCIENCE MONTHLY. 



When hatching begins, the water in the trough is raised, and 

 the trays are brought near the surface and held there by light 

 wedges which do not spring the sides of the troughs. The feather 

 sweeps the bottom of sediment, taking care not to injure the deli- 

 cate embryos, for they are more delicate when first hatched than 

 when in the egg. At this time cleaning should be confined to the 

 removal of any dead eggs or fry and of egg-shells on the screen, 

 and here is where the little flat net of milliner's " millinet," on 

 a three-inch brass-wire frame, is found most useful. This wire 

 is covered with muslin to which the netting is sewed, and, moved 

 quickly through the water, it gathers the shells which are clog- 

 ging the screen, and a smart rap on a pan cleans the net. It is 

 also of use in taking out dead fry and foreign matter from the 

 bottom, as well as deformed fish, for we find crooked-tailed fish 

 that can not swim and take food, double-headed fish, and twins 

 attached to one sac, which never live. 



When first hatched the embryo trout hardly look like fish ; 

 they have simply burst the shell in which they appeared as a 

 slim, dark body, with big eyes, coiled around the yolk, and now 

 merely straighten out and have the great yolk-sac still attached, 

 and so heavy that they can not swim, but lie on their sides and 

 huddle together to avoid light ; and now there is danger of the 

 bottom ones being smothered. Cover the trough in spots to in- 

 duce them to gather in different places, and keep them as dark as 

 possible. At this stage there is nothing to do but to remove the 

 few dead sediment can not hurt them now and keep the outlet 

 screen free. When they get to be about a month old the sac will 

 be nearly gone, and the fry begin to show signs of swimming by 

 occasional darts from the bottom to examine some floating par- 

 ticle. They will take food some days before the sac is absorbed, 

 and it should be offered to them in small quantities. 



The best food that I have found is beef liver, after abandon- 

 ing it for soft clams {My a arenaria) salt-water mussels, and horse 

 meat. The clams made the young fish grow fast, but did not pro- 

 duce the expected number of eggs ; the mussels were tried raw 

 and cooked, with the same result ; and, as the principal object at 

 this State hatchery was the production of fry for planting pub- 

 lic waters, we. next tried raw horse flesh, which was very objec- 

 tionable on account of the large proportion which was passed un- 

 digested and clogged the screens and fouled the water. Several 

 fish culturists have found an admixture of bran, shorts, or mill 

 feed with liver to be excellent. This I have not tried, because 

 the trout is not a vegetarian in the least degree, and it remains to 

 be proved that such vegetable addition to the food is of real ad- 

 vantage. 



The liver is fed raw ; for the " babies " it is run through a 



