PROPAGATION" AND DISTRIBUTION OF FOOD FISHES, 1921. 27 



dam to permit the passage of fish it was never effective. The matter 

 of providing a suitable fishway was taken up with the California 

 Fish and Game Commission, and at the close of the year the case was 

 pending in the courts. 



There was a very large run of fall chinooks in the Sacramento 

 Eiver, and while the early fall was marked by low-water stages it 

 was estimated on November 15 that there were enough fish impounded 

 at the Battle Creek substation to yield not le&s than 15,00(X()00 eggs. 

 At Mill Creek the run was even larger, and the egg collections for 

 that substation were estimated at not less than 30,000,000. Beginning 

 November 20 several days of heavy rains occurred, bringing about 

 flood conditions of unusual severity. Many miles of territory, in- 

 cluding railroads and highways, were under water. The retaining 

 racks at all points were destroyed and spawning operations abruptly 

 brought to a close. 



At the Battle Creek substation 2,450,000 eggs were taken, of which 

 504,000 were transferred to Baird for incubation and 20,000 shipped 

 to central station, Washington, D. C, for exhibition. The remainder 

 were hatched successfully, and 1.781,000 fingerling fish were planted 

 in Battle Creek, At the Mill Creek substation 5.400.000 eggs were 

 secured, of which 1.000.000 were shipped to Baird and 3,000.000 to 

 the State hatchery at Sisson. From the remainder 1,347,400 finger- 

 ling fish were produced at the station. 



INVESTIGATIONS AND EXPERIMENTS, 

 FEBTILIZATION OF SALMON EGGS IN NATXJBAL SPAWNING. 



At many points on the Pacific coast certain species of salmon, 

 notably the humpback and chum, spawn at or near the mouths or 

 streams in tidewater. A rather general belief has existed that the 

 results of such natural spawning were negative, and in order to ob- 

 tain information on the sul)jcct the field superintendent in charge of 

 the Pacific coast stations and the superintendent of the Baker Lake 

 (Wash.) station were detailed to investigate these conditions in 

 Puget Sound. The report of their findings follows: 



In conformity with in.stnictions relative to securing information on the per- 

 centage of fertilization of salmon eggs in natural s(»awning, with special refer- 

 once to the eggs deposited in places over whicii the tide ebhs and flows, an 

 investigation was made of Walcott Slough, at Brinnon, Wash., on March 5, 

 1920. 



Walcott Slough runs through extensive tide flats, which at flood tide are 

 completely coveretl with salt water. During ebb tide a good flow of fresh 

 water, estimated at about 3,000 gallons per minute, passes through the slough, 

 its source being spring seepage at the slough head. Tiie rtsh — chum salmon 

 witli a few silver salmon — enter this slough to spawn, the entire slough bottom 

 being composed of gravel suitable for the purpose. The bureau has a rack and 

 frai» installed at a favorable point in the slough, about r)00 yards from salt 

 wntfT. where flsh are captured for prf)pagation. Helow the rack some natural 

 spawning occurs, and this occupied our attention. Several nests were in- 

 ve.stigatefl, all but two of which contained eggs. These two might possibly 

 have been made last year, though they had the appearance of being used quite 

 recently. Eight nests were examined, l»egiiining with Xo. 1 (see table, p. 28) 

 Just below the trap, about .^»tK) yards from salt water at low ti<le, and ending 

 with .No. 8, about 20<^) yards from the same water stage. During flood-tide 

 periods wilt water covered the nests from three to six and three-fourths liours 

 to a flejttli of li to ui feet. This will vary somewhat with higher or lower 

 runs of tide. 



The condition and character of the bottom wherever nests were found would 

 indicate that the flsh chose points where spring seepage occurs through 



