FISHERIES OF THE PACIFIC COAST. 
213 
possible, and extending the bottoms inward to connect with inclined troughs placed 
along the axis of the wheel, so that when it revolved, gradually increasing the angle 
of the bottom of the net with the horizontal, the fish would gravitate towards the 
trough and slide out into a box placed below the shore end of the axis of the wheel, 
and we have a device for catching fish with a dip net that would appear simply 
divine in the eyes of the untutored savages who wielded it so long under the condi- 
tions of this locality. 
The success of this fishing machine depends on the fact that there are many points 
along the rapids of the river where the trails of the traveling salmon come together 
alongside of rocky points projecting from the shore. A wheel placed so as to cover 
such a trail will be tolerably certain to catch the majority of the passing fish. There are 
only a few points where these conditions are maintained constantly during the varying 
stages of the river, and where wheels may permanently be placed, and hence some 
wheels are placed on the ends of scows, which can be moved from point to point to 
meet the varying conditions of river stage. The success of the fish wheels is painful. 
It has been aptly remarked that they pump the fish out of a river. And yet, if it be 
true that nearly all of the fish die after spawning, I can see no objection whatever to 
their use, provided they be required to stop fishing for two days, say Saturday and 
Sunday of each week, and that at the same time at least one hatchery be established 
on the upper waters of the river. 
It has been demonstrated that a hatchery can insure the placing of enormous num- 
bers of young fish in the sea annually, and the evidence of Mr. G. W. Williams shows 
conclusively that the place for propagating the Chinook salmon is at their spawning 
grounds on Snake River. The wheel at the Cascades of the Columbia River, owned 
by Mr. Williams, catches more fish than any other. The maximum catch reported 
was 50,000 pounds in one day. At the time. I visited the locality it was catching at 
the rate of 20,000 pounds per day. These wheels can only operate at certain stages 
of the river. At the Cascades they can not operate when the water is less than 15 
feet above low water, nor when it is more than 47£ feet above. 
While at the Cascades and The Dalles it was stated to me that 90 per cent, of the 
fish caught at these points were bluebacks and steelheads, and the catch during my 
visit verified the statement. This shows that about nine-tenths of the large Chinook 
salmon are taken by the nets in and-about the mouth of the Columbia River. Unless 
something be done to allow a larger percentage than this to pass up the river it will 
hardly be possible to get enough of this variety on the upper waters for propagation. 
The fish wheel is a patented device. It was first used by the pat- 
entees, Messrs. G. W. Williams & Brother, in 1879, and for three years 
those gentlemen had a monopoly of the fishery with this form of appa- 
ratus. About 1882-83 other parties built fish wheels, and fishing in 
this manner has rapidly increased since that time. In 1889 there were 
27 stationary wheels and 18 scow wheels on the Columbia, from the 
vicinity of the Cascades to Celilo. Reference is made to the map (plate 
xxxii) for the location of these. 
Trawl lines are used to a limited extent for the capture of sturgeon. 
Some of these are of the Chinese pattern, fitted with barbless hooks 
placed 15 inches apart, with 100 hooks to each trawl. Others are rigged 
with ordinary American-made barbed hooks, upon which bait is used. 
Methods of fishing , catch , etc . — Gill net drift fishing is prosecuted 
chieflyjm the estuary of the Columbia in and near the channel. If the 
water is clear, the nets are set only at night ; but a rise in the river, 
which is favorable for the salmon to “run,” makes the water muddy, 
and at such times fishing is carried on both day and night. Night fish- 
