784 BULLETIN OF THE BUREAU OF FISHERIES. 



5. That the method proposed is superior in that it follows natural con- 

 ditions governing the propagation and welfare of the fish, only eliminating and 

 providing against the destructive forces, such as floods, drought, the tendency of 

 trout to prey upon the eggs and young, and protection against such other fish 

 and animals as prey upon the eggs, embryos, and young fish. 



6. That the proposed method of causing the trouts (and probably under 

 favoring circumstances, most of the salmon) to deposit their spawn in prepared 

 runs or races, and the subsequent care of the nests and the young fish, may be 

 more economically carried out than the artificial method of collecting eggs, 

 impregnating them, and thereafter caring for them, as it is now practiced in 

 most hatcheries, involving expensive plants and skilled attendants. 



REPORT OF EXPERIMENTS. 



In support of the foregoing claims for the advantages of the system outlined, 

 it is manifestly impossible to submit a portable model or other more tangible 

 evidence than the sketches and particulars herewith submitted. The facts of 

 experiments made are briefly summarized as follows, with the aid of diagrams 

 I, 2, and 3: 



With a series of four ponds, constructed within a few yards of Boulder 

 Creek, in which the rainbow trout are native, water sufficient to provide a flow 

 through the ponds was diverted thereinto by way of an open trench 300 feet 

 in length. The ponds are about 30 by 60 feet in area and range from 2 to 6 

 feet in depth. Several falls over weirs aerate the water suftlciently. The em- 

 bankments are walled with bowlders, laid up without masonry, and in all respects 

 the ponds comply with the natural conditions of the stream as nearly as can be 

 devised. The temperature of the water in the ponds ranges from 40° to a maxi- 

 mum of 83° F., the latter high temperature occurring several days in the month 

 of August, 190S, and lasting but a few hours in the afternoons of the warmest 

 days. The fish suffered no ill effects from this extreme temperature, but were 

 for the time manifestly restless and alarmed. 



For two years, covering the spawning seasons of 1906 and 1907, from 40 to 

 100 adult rainbow trout were held in the ponds. These trout were taken from 

 Boulder Creek with hook and line, readilv became domesticated, all remained in 

 good condition, and are at present among the largest of the breeding fish in 

 the ponds. 



The first season (October, 1905, to April, 1906), the larger of the fish 

 spawned in beds made around the shores of the ponds, and in due time between 

 100 and 200 fry reached the surface. The little fish, with the exception of half 

 a dozen, disappeared within a month. Five or six only survived. 



The second year (1907), while a larger number of the parent fish spawned 

 still fewer fry appeared, and but four of these reached the yearling stage. This 



