No. 3. The inclosure in use for the confinement of the stock of breeding fish for the four years from 1872 to 1875, inclusive, was made by running a barrier across a narrow arm of a small lake (mentioned in official reports as "Spofford's Pond") near Bucksport village. This body of water, about 60 acres in area in the summer, receives the drainage of not more than 5 square miles of territory through several small brooks, that are reduced to dry beds by an ordinary drought. About a quarter of the shores are marshy and the rest stony. The water is highly colored by peaty matters in solution, and all objects are invisible at a depth of 2 feet: The bottom is composed mostly of a fine brown peaty mud of unknown depth. Aquatic vegetation of the genera, _Nuphar_, _Nymphaea_, _Bragenia_, _Potamogeton_, &c., is abundant. The water is nowhere more than 16 feet deep in the spring, and 11 feet in midsummer. The portion inclosed is 2 feet shoaler. The inclosure occupied sometimes 8 or 10 acres, and sometimes less. The barrier was from 400 to 600 feet long, and was formed the first year of brush; the second and third years of stake-nets, weighted down at the bottom with chains; and the fourth year of wooden racks, 4 feet wide and long enough to reach the bottom, which were pushed down side by side. The brush was unsatisfactory. There were holes in it by which the fish escaped. A single net would not retain its strength through a whole season, the bottom rotting away and letting the fish out, unless before the autumn was far advanced its position were reversed, the stronger part that had been above water being placed now at the bottom. This method was therefore rather expensive and not perfectly secure. The wooden racks were costly and heavy to handle, but quite secure. The salmon placed in this inclosure had to be carted in tanks of water overland about a mile in addition to transportation in floating cars from 3 to 5 miles; they were transferred suddenly from the salt water of the river (about two-thirds as salt as common sea-water) into the entirely fresh water of the lake. To all the supposed unfavorable circumstances must be added the high summer temperature of the water. During August the mean was generally above 70 degrees Fahrenheit at the bottom and several degrees warmer at the surface. Occasionally there was observed a midday temperature of 74 degrees F. and once 75 degrees at the bottom. Yet this proved an excellent place for our purpose, a satisfactory percentage of the salmon remaining in perfect health from June to November.