FISH-CULTURE. 175 



when wet soon bleaches, and then the fish will not touch it ; 

 neither do the fry care to pick any motionless particle off the 

 bottom. The solution which occurred to me was to enclose finely 

 grated liver in a tube, and cover the end of the tube with coarse 

 wire net, a piston at the other end slowly forcing the meat 

 through the net. So I made a small water-wheel, with buckets set 

 on its circumference, and the axle bored to receive the thread of a 

 male screw. This screw was cut on a shaft or piston-rod, a part of 

 whose length was square in section, and ran in a guide. Water 

 dropped into one bucket at a time, and, as it filled, the wheel 

 turned a small part of a revolution, forcing the piston forward. 

 The supply of water was reduced to fill the bucket two or three 

 times per minute, and each time the bucket filled and emptied, a 

 small portion of food was forced through the netting and fell into 

 the rearing-box or pond. 



For experimental work this device is useful. It works much 

 better with dry food than with liver, but it is absolutely valueless 

 for work on a large scale. Whether any other than hand-feeding 

 is advisable where the number of fry is so great as to occupy 

 the whole time of one person is a moot point. But there will 

 always be experiments and odd lots out of the ordinary routine, 

 and with these mechanical feeding may be a convenience. There 

 are other and better plans, which will be described in their 

 place. The best and simplest is to form the food into thread 

 resembling maccaroni ; the movements of the fry and the swirl of 

 the current keep these artificial worms always en evidence. There 

 is no waste ; the food never gets time to blanch, and the surface 

 exposed to the water is small. 



I still hoped to find a mechanical feeder, one that was handy 

 and cheap, so that it could be used in all the rearing-boxes. My 

 next plan was to place the food in a glass jar, fitted with a neck in 

 the centre of the side. In this I placed a cork, through which a 

 bent tube passed nearly to the bottom of the jar ; the top of the 

 jar was covered, and a straight tube passed through the cover, 

 ending in a funnel. Water dropped slowly into the funnel, fed by a 

 syphon from the hatching-house aqueduct. The food was kept in 



