798 BULLETIN OF THE BUREAU OF FISHERIES. 



accustomed to it. It is a cheap and generally available staple. Food prepared 

 as described may be readily dried and preserved for emergencies where a fresh 

 supply of meat is lacking. 



That millions of trout and salmon fry have been and are being maintained 

 in overcrowded hatching troughs upon a diet of beef liver would appear to 

 be positive evidence of its great value, while it is commonly as easily and cheaply 

 obtainable as any form of animal food. 



The chief object of this paper, however, is to suggest that young salmons 

 and salmonoids reared in captivity should be given the minimum quantity of 

 artificial food and a maximum area and flow of water containing their natural 

 food, for which they should be permitted to forage. Prepared food should 

 supplement the natural supply where water area is overcrowded with young fish, 

 or where drouth, cold, or other climatic conditions interfere with the normal 

 natural supply. In support of this view is offered the following summary of 

 well-known or readily ascertained facts and examples: 



1 . That along the salmon rivers and trout streams fry existing under natural 

 provisions are commonly in excellent physical condition, mortality among them 

 being mainly caused by abnormal disturbances of the nests, such as floods, 

 drouths, or extraordinary climatic changes, or by the depredations of natural 

 enemies, birds, reptiles, and other animals. 



2. That salmonoids are not surface-feeding fishes exclusively, but seek 

 food suspended in the water and on the shores and bottom surfaces accessible 

 to them ; and that of necessity they must collect more or less vegetable and sedi- 

 mentary matter; in fact, that they are rather omnivorous than piscivorous or 

 carnivorous fishes. 



3. That under normal natural conditions a continuous succession of season- 

 able aquatic and insectiverous foods, much of which will embrace vegetable 

 matter in some form, is supplied to the young fish. 



4. That owing to the minute particles of food matter collected by newly 

 hatched salmonoids, it is doubtless impossible to distinguish with accuracy the 

 natural or instinctive selections made by them, or to determine nutritive values. 



5. That it will appear that suitable natural food for salmonoids is abimdant 

 in the waters wherever trout and salmon spawn, and that the most available, 

 economical, and scientific provision for young salmonoids may be made in the 

 preparation and adaptation of sufficient water area in normal natural condition, 

 but subject to control as regards floods, drouths, freezing to extremes, and the 

 exclusion of destructive animals. Controlled areas of stream or prepared runs 

 should provide for the absolute regulation of the water flow, and should contain 

 trap pools or other devices for collecting the fish, excluding them at the end of 

 the spawning season, and finally reducing the flow of water to a minimum for 

 the purpose of capturing the fry or young as may be desired. 



