MARINE FISH HATCHERY AT FLODEVIG, NORWAY. 



805 



back end as well as both sides of the collector are covered with silk gauze no. 40. 

 All water coming from the pond will thus have to flow into the collector, and as 

 this will act as a strainer the eggs will be kept back, while the water continues 

 its course through the gauze netting toward the overflow pipe g. As the 

 opening in the wall as well as in the chute is deep and wide, the outflowing cur- 

 rent would be too slow to bring all the eggs in the pond into the collector in a 



/ ' /'I e ;■ -' ' -' 



;. i: ■; r' "' 



' I' < ■ ' 



' '! I , ' 



/ - :l I ' ; 



js^ 



Fig. 4. — Device for installation of egg collector. (Plan.) 



reasonable time. To remedy this, a partition or dam has been placed at h 

 (fig. 4 and 5) , and at the same height as the upper end of the overflow pipe. 

 Instead of a slow current 18 inches deep, we will now have a strong current 

 half an inch deep, and as the eggs float at or near the surface of the water, 

 all of them will in a short time, say two or three hours, be drawn into the collector. 



Fig. 5. — Same as figure 4, showing section A-B. 



I have mentioned above that the cod spawn in the evening. Consequently 

 all the eggs would be in the collector at about midnight, and have to remain 

 there crowded together till the men arrive in the morning. As this is not 

 desirable, the surface outflow is stopped in the evening and the wooden pipe 

 (flg. 2), which draws the water from a greater depth, is opened, and thus the 

 eggs will remain quietly at the surface of the pond till morning when the sur- 

 face current is tvuned on again by closing the wooden pipe. 



