78 
of the United States Fish Commission, and also at the 
Troutdale Fish Farm, at Mammoth Spring, Ark. The 
method followed at these three places is not, so far as I 
know, prosecuted elsewhere; the differentiation consist- 
ing in an admixture of vegetable matter with the flesh or 
animal matter, heretofore constituting the sole food for 
trout under domestication. A few notes on the methods 
of preparing and administering this food at the Neosho 
Station will illustrate the method of the three places 
where it is used. A thick mush is made by cooking 
‘‘shorts’’ or mill-middlings in boiling water, which, after 
it has thoroughly cooled and stiffened, is mixed with 
liver ground to a fineness suitable to the size of the trout 
tobefed. Thevery young trout have never been subjected 
to this diet (though it is not doubted that they could be 
induced to eat it), but they are started and kept upon a 
pure beef liver diet until they are thoroughly trained to 
congregate for their food. When the fry have been on 
liver for about two months, we commence to mix in a little 
mush, and gradually increase the proportion of mush (and 
quantity of food) until by the time they are six months 
old the mush and liver are in equal proportions. After 
that time the addition of mush is made freely, so that 
when the fish are yearlings the liver may be reduced to a 
minimum. 
Exigencies have arisen, making it desirable to economize 
on liver. At such times we have not hesitated to put the 
trout on a diet of pure mush. They do not allow this food 
to sink to the bottom and eat it only when pressed by 
hunger. On the contrary, they rise to the surface, some- 
times eat it in the air, and rarely or ever allow a particle 
to reach the bottom. To 1,000 yearling trout we have been 
giving adaily average of 1.87 lbs. of this mixture, in the 
proportion of 3.79 mush to 1.0 liver. Their average length 
at one year old was 6in., and the weight for an average 
1,000 was 51.86lbs. The loss in raising 40,000 trout to 
