FISHERIES OF WASHINGTON AND BRITISH COLUMBIA. 297 



as little reliance cau be j)laced upon the existing knowledge of the 

 movements of the fish. How the growth of the industry may affect 

 operations on the Fraser River and the abundance of the sockeye is 

 also an important matter which remains to be determined. 



CONSTRUCTION OF THE TRAPS. 



The salmon trap nets are constructed on the same general principle 

 as the pound nets of the Great Lakes, consisting of a crib, tunnel, 

 heart, and leader; but they are usually made of a larger size, and 

 exi3erience has dictated some important modifications. The netting 

 is of cotton twine, and is supported by wooden stakes driven into the 

 bottom. Wire netting of galvanized iron, in place of the cotton, for 

 the hearts and leaders, has been suggested as probably more durable, 

 and experiments regarding it have recently been carried on at Point 

 Roberts. Floating traps, su<;h as are successfully employed for salmon 

 and other species in the Gulf of St. Lawrence, have never been tried 

 in this region, but their relative cheapness and the ease with which 

 they can be shifted from place to place are advantages which might 

 commend them to the fishermen of Washington. 



The fishing-sites in the track of the sockeye are largely in exposed 

 positions, many of them being open to the full force of any gale 

 sweeping across a wide exi>anse of water from more than one direc- 

 tion, as is especially the case at Point Roberts. This condition neces- 

 sitates the building generally of stronger traps than are customarily 

 used in other regions. The stakes are unusually heavy and are often 

 backed by additional piling. The crib, moreover, is frequently 

 strengthened by a capping of timber which binds the stakes together, 

 and this capping may be continued along the top of the heart and 

 even of the leader to a greater or less distance. This construction 

 gives the appearance of great permanency, but it is designed only to 

 meet the requirements of a single season, and it sometimes fails even in 

 this respect, especially if the season be a stormy one. While some of 

 the upper timbers and the netting may be saved, the stakes are seldom, 

 if ever, available for a second season. The latter are rapidly honey- 

 combed by ship-worms and it is not the practice to remove them. 

 They are liable to break in the attempt to pull them from the bottom, 

 and in the course of two or three months they become so thickly cov- 

 ered with barnacles as to chafe the nets badly. 



The length of the leader varies according to location and the slope of 

 the bottom, but it is generally much greater than in the Great Lakes, 

 sometimes exceeding a half mile. The cribs are also generally of extra 

 size, rectangular, but not always square in shape, and measured in the 

 several traps examined from 35 to 80 feet on a side. Their depth 

 ranged from 3 to 9 fathoms, dependent upon the depth of water. The 

 hearts are, as a rule, proportionally large for the size of the crib, are 

 sometimes double, one leading into the other, and constitute the most 

 novel feature of the trap. They vary greatly in shape to meet the sup- 

 posed exigencies of each locality, and often have a leader-like exteu- 



