spicuously lack shelter. Rocky shores, stumps, or other shelter facilities are normally present. The most frequent type of stream improvement is construction of pools in streams which are too shallow, with long areas of flat water where even the most optimistic angler would not expect a good-sized trout to be living. Temperature protection, through narrowing exposed areas by use of deflectors or by lowering the channel by digging is another important aspect of the work. Cleaning and ditching of some of the marshy spring runs is another practice which aids temperature conditions. The planting of willow cuttings, done at three camps last spring on an experimental scale, gave excellent results. Especially on the wide, warm, flat streams which are so prevalent in the southern and central part of the State, planting of willows is a method of environmental improvement which should be of increasing importance in the future. The flood of last summer, which struck the smaller streams of central New York with unprecedented severity, served to emphasize the hazardous condition of many streams. When trees are present the stream banks are fortified by the roots and the water is forced toward midchannel. This results in scouring action of midchannel and streams run deep and relatively narrow, keeping their channels clear of debris which would otherwise block them. Clearing of streamside brush is a common practice, since each farmer usually wants to increase the area of productive pasturage or crop lands. Removal of the alders, willows, or other trees allows bank erosion and as the banks give way the streams become wider. Since the transporting power of the water on this wide frontage. becomes lessened, the channel tends to block with stones and the result is a wide, flat stream. Temperature and pool conditions are usually unfavorable for trout. This is not the whole story, however, since the hazard of a stream of this wide, shallow type is not confined to damage to fish. Flood waters, no longer forced into a midchannel path of maximum erosion, erode at the edges, skinning off good agricultural land. It would pay many landowners, in the long run, to leave a zone of brush and trees or to plant them, where they do not grow naturally. This seems to be the most economical way to get the streams back into less hazardous condition. Erosion is being controlled as a part of the trout-stream improvement work. This is attacked by turning streams away from points where destructive erosion is developing or is likely, by removing debris where this is located so as to turn the current toward the banks and by building log retaining walls to strengthen banks. Some large slides, which annually contribute tons of sand or silt to streams, have been checked. Another side of this work consists in protection of trout streams against undesirable fish which run up from below. A barrier dam, with log apron so arranged that perch, pickerel, or other fish cannot jump, has proven a very successful construction in preventing upstream migration of undesirable fish. Split streams, which run in several channels and do not have sufficient water in any one channel to form good pools, and con trolled by blocking side channels in order to get all the water together. Improvements in the character of the water have sometimes been made by decreasing the amount of water standing on decomposing muck and by keeping the stream moving along more rapidly. Small dams, faced with boards, readjustment of boulders to form rocky pools, deflectors which turn the water, and many other constructions are useful. It is the aim to do this work economically, using logs and boulders available at the stream, and to put the stream in as good condition as possible with the means at hand. Elaborate work is neither justified nor desirable, and no one likes to see artificial-looking work along trout streams. Permanence is an essential, of course. There is reason to believe that properly constructed improvements will last many years. Inspection of two improved streams hit by the flood of last July indicated total damage not exceeding 10 percent, which is remarkable considering the extreme violence of this flood. Most of the modifications are along lines for which Nature sets the pattern. Usually, in looking over a stream, a good working basis can be derived from study of the best of the natural places. In checking the results of stream-improvement work, there is an important handicap in that it has not proven practicable to study each of the many streams intensively for a long period before and after the work. In checking over a number of samples of the work done last year, I have been well satisfied that more good-sized trout can be carried than before improvement. I do not mean that any and all improved streams are fishermen's paradises. A small stream, without good pools, produces only a few fish of legal size. Improve it and it produces more. It still remains a small stream, easily fished out of large fish. Fishing pressure acts to keep large fish scarce on improved as well as unimproved streams. The degree of success would need to be measured by catch records before and after improvement, not by measuring what is left in the stream. Unfortunately, catch records on each of the streams are at present impossible to obtain. It may seem surprising that catches of trout are made a year after improvements, in areas formerly unsuited to production of legal-sized fish. Most streams, if any good at all for trout, have quite a number of young fish toward their headwaters. Such young fish soon take possession of improved areas and after growing 1 year are likely to be legal-sized fish unless hindered by excessively cold water or poor food supply, such as is induced by excessive crowding. The management of unfavorably balanced populations in lakes and streams is a subject requiring much investigation of principles and applied methods. In fact, such work as has been done is to be regarded as more or less in the nature of pioneering. The growth-rate investigations of the Biological Survey have disclosed a great range of rapidity of growth of trout, bass, perch, and some other species. We have lakes where perch run large and lakes where they are small and also very numerous. Our survey reports have pointed out that lakes frequently develop large populations of such fish as perch, sunfish, and rock bass, crowded, stunted, and of slight attraction to the angler but keeping the lake 57282-36- -29 in a state of poor production. You might say, well then, plant heavily with young bass or other species to forage down these others. But where we have a crowded population of 3- to 5-inch rock bass, some of which are over 5 years old, it is hardly logical to expect that stocked fish can increase against this competition. The only parallel that I can think of would be for a farmer to sow his corn down in his scrub pasture or woodlot instead of in carefully prepared ground. Even on plowed land he gets no crop unless he cultivates to remove weeds. When we do progress sufficiently, and it will take time and money, of course, in the field of removing excess fish the technique should be patterned after the cultivation of field crops in that it will be a problem of controlling what is locally undesirable rather than giving any species a black eye and trying to exterminate it in all waters. Some of the principal unfavorable environmental conditions that the surveys have brought to light as more or less prevalent in the waters of the State can be listed as follows: (1) High temperatures (affecting trout especially): To some extent subject to control through ditching of spring, narrowing of streams, and planting of shade. (2) Poor pool conditions, uniformly shallow areas where fish find little shelter, especially in low water: Most of the stream improvement done has been devoted toward remedy of such conditions. (3) Unbalanced population conditions: Particularly prevalent is overcrowding of relatively less valuable forms of fish life, and overcrowding and stunting of game fish. Some species are undesirable in the particular waters they inhabit and interfere with species which would otherwise do well in these waters. Competition is often too severe to allow each individual the food supply that it should have in order to produce a fish of valuable size. Barrier dams have been used to keep undesirable fish, particularly perch and pickerel (northern pike) from migrating up into trout waters. (4) Depletion of the more sought-after species: This tends to upset the balance already mentioned, by allowing unfavorable species to increase. This is difficult to control. The work done has increased water areas and probably added slightly to chances for escape of fish. (5) Parasites: Particularly important here are the grubs which affect the value of fish to the angler and internal parasites which affect the health of the fish. Possibly other diseases than those caused by parasites should be listed, although there is as yet little information available on the seriousness of most diseases in wild waters. Decrease of ponded areas may have a limiting effect on pond snails, the intermediate hosts of grubs. Work done, so far, has probably had slight effect on parasite conditions. (6) Unfavorable spawning places: Dams, both those built by humans and beavers, should be listed here with respect to their action under some circumstances in blocking fish from reaching their spawning grounds. This factor is supposed to have been responsible for the extermination of the natural run of the Atlantic salmon, for example. Wherever it has been practicable to remove barriers, clean out spring runs, and increase gravel beds, spawning areas have been favored. It is doubtful if lack of spawning areas constitutes a prin |