66 EEPOET OF THE COMMISSIONER OF FISHERIES. 



spawTimg grounds. Over four-fifths of these were taken in tlie vi- 

 cinity of Sodus Point, N. Y., and more might have been secured there 

 had hatching space for handhng them been available. 



In connection with the propagation of compiercial fishes, the 

 usual numbers of young brook and rainbow trouts were hatched at 

 Cape Vincent, the eggs having been furnished from outside sources. 



In advance of the spawning of the pike perch in April, trap nets 

 for the capture of brook fish were set in the Oswegatchie River 

 within the corporate limits of Ogdensburg, N. Y., with the intention 

 of conducting spawning operations in cooperation with the New 

 York fisheries authorities. However, for some unknown reason, the 

 catch of fish was only about one-third as large as last year, when 

 operations were conducted on the same stream. This resulted in a 

 proportionate decrease in egg collections, the total deliveries at the 

 hatchery amounting to only 21,312,500, or less than one-third the 

 number obtained in 1916. The fry hatched numbered 10,875,000, 

 and after providing for a proper return to the spawning grounds the 

 remainder were shipped to applicants in New York and Pennsylvania. 



During the spring of 1917, 31,350,000 yellow-perch eggs were 

 collected from brood fish obtained by setting nets m the river near 

 the hatchery and 26,000,000 were hatched. With the exception of 

 250,000, which were used to fill applications, all of the fry resulting 

 from this work were returned to the river. 



The experience of recent years having demonstrated that pike- 

 perch propagation could be more successfully and economically con- 

 ducted on the shores of Lake Chajnplain than at the established 

 location on the Missisquoi River, steps were taken early in the fiscal 

 year to take down the Swanton hatchery and rebuild it on a selected 

 site at Sandy Point, on Missisquoi Bay. All of the old material was 

 utilized in the new construction, which w^s erected on a cement 

 fomidation and provided with a cement floor into which two large 

 fry receiving tanks were built. The hatching capacity of the new 

 building w^as increased by about 100 jars, and an abundant water 

 supply by pumping was arranged for by extending the suction pipes 

 directly into the lake. As funds were not available for completing 

 the hatchery in all its details, only such part of the construction 

 was planned for as would permit of the propagation of pike perch 

 during the spring, with the view of adding the finishing touches 

 later on. 



The cold, backward spring and the presence of ice in the lake and 

 river about one month later than usual were mainly responsible for 

 the poorest spawning season that has been experienced in this region 

 for some years. Such conditions have always been found to seriously 

 curtail the run of spawning pike perch, tliis being especially notice- 

 able in 1913, when practically no females entered the river. That 

 year, however, they were taken from the lake, but during the past 

 spring the fish did not seem to congregate on the lake spawning 

 grounds at all. 



In the belief that the old method of securing a brood stock from 

 commercial fishermen entailed more expense than necessary, the 

 station superintendent engaged the services of an expert web worker 

 from the Lake Erie field for the construction of a trap net, to be 

 operated by the station force. After a trial it was decided that the 

 net could not be worked in the river successfully owing to its frequent 



