No. 4. The inclosure in use since 1870 at Dead Brook, Bucksport. It is located in a gently running stream bordered by marshy ground, with a bottom in part of gravel but mostly of mud, crowded with aquatic vegetation. The water, supplied by two small lakes among the hills, is cleaner than the average of Maine rivers, but does not in that respect approach the water of inclosure No. 1. The greatest depth is about 8 feet, but in the greater part of the inclosure it is from 3 to 5 feet. The width of the stream is from 2 to 4 rods, and the portion inclosed is 2,200 feet long. The barriers to retain the fish are in the form of wooden gratings, with facilities for speedily clearing them of debris brought down by the stream. Better results were expected from this inclosure than from No. 3, but have not been realized. The percentage of salmon dying in confinement has been greater, amounting commonly to about 25 percent of those introduced, and this notwithstanding the salmon are conveyed to the inclosure by water carriage the entire distance (7 miles) instead of being carted in tanks. The cause of the trouble has not yet been discovered, but there is good reason for thinking that it lies in some of the circumstances attending the transfer of the fish from the place of capture, and that the inclosure itself is perfectly suited to its purpose. This view is supported by the fact that nearly all the losses occur within a few weeks after the introduction of the salmon and almost wholly cease by the end of July. If the cause of disease was located in the inclosure, we should expect it to be more fatal after a long than a short duration of the exposure of the fish to its action, and that with the smaller volume and higher temperature of August it would be more active than in June and July. The above description will, I think, give Dr. Bottemanne a sufficiently correct idea of the character of the inclosures we have tried. There are, however, several other points to be touched upon to put him in possession of the practical results of our experience. The facilities for the recapture of the salmon when the spawning season approaches must be considered. In the lake at Bucksport village (No. 3) we hoped at first that their desire to reach a suitable spawning ground would induce them all to enter the small brook that forms the outlet, which was within the limits of the inclosure. In this matter our expectations were but partially realized. Many of the fish refused to leave the lake through the narrow opening that was afforded them, and were only obtained by pound-nets, seines, and gill-nets, all of which involved a considerable expenditure of labor and material.