164 FISH CULTUEE 



results. One is liver, prepared by rubbing it 

 through perforated tins so as to reduce it to 

 the consistency of a pulp or paste ; the other is 

 thick milk. The liver paste should be mixed 

 with water untU it is thin enough to be drawn 

 through the automatic feeder drop by drop, at 

 such a rate as will enable the fish to feed 

 steadily from an hour to an hour and a half be- 

 fore the reservoir is exhausted. If it is made 

 too thin, it will be dropped into the ponds faster 

 than it can be eaten. When the feeding is to be 

 done by hand the liver paste may be made a 

 little thinner than when prepared for the auto- 

 matic apparatus. 



In the preparation of milk-food the best re- 

 sults are obtained by using skimmed milk which 

 has been allowed to curdle and thicken natu- 

 rally. The thickening ought not to be forced by 

 scalding or boUihg, which gives a toughness to 

 the curds that is undesirable, and which is not 

 so easily digested as the soft particles of nat- 

 urally thickened milk. It should be placed in 

 the automatic feeders in the same quantities 

 as the liver. A little salt, occasionally, in both 

 the milk and liver paste, has been found to be 



