ATKINS ON FISH-WAYS. 597 



an iuvitation, after securing' a proper location, is to pour out a fair flood 

 of water from the lower end of the lish-way. As has already been re- 

 marked, the migratory tislies prefer a large stream rather than a small 

 one. Though any stream, issuing from the obstruction or near it, pos- 

 sesses the capacity" to attract them to a certain extent, the attraction of 

 the smaller is wholly or partially overcome by that of the larger stream. 

 A fish-way may discharge a volume so small as to escape for many days 

 the notice of the fish for whose use it was designed. So far as can be judged 

 from its observed conduct, a fish is guided in its ascent of a river almost 

 solely by the current, knowing little more than that its way is upward. 

 When it meets the obstruction of a dam, there are several distinct cur- 

 rents beckoning it on ; there are the currents from the waste-way, the mill- 

 wheels, and the leaks, all competing with the current of the fish-way. 

 It commonly chooses first the largest stream, and not learning rapidly 

 from experience, nor early losing courage under repeated and constant 

 failure, it may pass days in repeated ineffectual struggles against the 

 impassable flood of the main stream, alternating with unsuccessful 

 searches for an easier way. It is quite evident that the larger the vol- 

 ume flowing from the fish-way the sooner will the fish find and enter it. 

 It is a reasonable proposition that this volume should bear some defi- 

 nite proportion to the total volume of the river, or to that part of it 

 which passes down bj' the same channel into which the fish-way emp- 

 ties. The only attempt to fix such a ratio of which I am informed was 

 made by Mr. Samuel U. Roberts in 1809. In discussing the construction 

 of salmon-ladders in Ireland, Mr. Roberts laid down the following rules : 



"Theleast quantity which will supply asalmon-ladderefficieatly is that 

 which will pass through an opening 8 inches wide, with a depth of 18 

 inches of water flowing through it, and the maximum quantity need not 

 exceed that which will flow through an opening 3 feet wide with the same 

 depth. 



" I am of opinion that between these two limits a scale may be formed 

 which will secure an adequate supply for salmon-ladders. It occurs to 

 me that the fair basis on which to regulate the quantity is the extent of 

 the rain-hasin of the river, and I am of opinion that on all new weirs 

 constructed on rivers, whether for navigation or mill-power purposes, 

 the minimum opening to be provided should not be less than 8 inches in 

 width, the sill being 18 inches below the apex or top of the weir ; 

 that the depth should be fixed in all cases at IS inches, but that 

 the width should increase in the following proportions: '" Wheu 

 the rain basin exceeds 50 square miles, the width to be increased at 

 the rate of 1 inch for every additional 50 square miles until it amounts 

 to 12 inches; when it exceeds 250 square miles, the width to bo in. 

 creased at the rate of 1 inch for every additional 100 square miles until 

 it amounts to 18 inches. When it exceeds 850 square miles, the width 

 to ba increased at the rate of 1 iuch for every additional 200 square 



