a different time and would not be likely to frequent the spawning places at the same time as the trout. I have no doubt that an infinitely greater proportion of trout ova are eaten by the trout themselves than by grayling in rivers which contain both fish. Chalk streams and those rivers with gravelly bottoms and with alternate shallows and pools seem to be the most suited to the grayling.

Among coarse fish the rudd is one of the best from the fly-fisher's point of view. It takes the fly readily, is very prolific and very easy to introduce. It thrives remarkably well in ponds which contain a good supply of food. Its fry serve as excellent food for other fish, particularly trout, but I have known cases where it increased rapidly in a pond at the expense of the trout. It can, however, be kept under by judicious netting.

The dace is another fish which gives sport to the fly-fisherman. It will not thrive in ponds. In some rivers, however, where trout—brown trout, at any rate—will not thrive, the dace does very well. In the case of the Sussex Ouse this is most remarkable. Little more than ten years ago there were no dace in that river, now it swarms with them. Their presence is attributed to the fact that some dace, brought there as live-baits for pike, escaped destruction and established the present stock. Sluggish and muddy rivers seem to produce the best dace. Chubb, which also possess many points to recommend them to the fisherman, will also do well in such rivers.

To those who enjoy bottom fishing and possess a pond, even a small one, I can recommend no fish more highly than the king-carp. It is a much bolder-feeding and gamer fish than the common carp, and is just as easy to introduce. While dealing with carp I may mention that the goldfish, when introduced into a suitable pond, grows to a very large size. I have caught them over a pound in weight.

The perch is a very prolific fish, and will thrive in ponds with a very small stream running into them, and in sluggish rivers. Other coarse fish are as a rule easy to introduce into a water. Though perch fry form excellent food for trout, perch, and of course pike, should be kept out of a trout water.

The suitability of a water depends to a great extent (as to its capacity of supporting a healthy stock of fish) upon its having plenty of suitable vegetation upon the banks. Therefore if the banks are bare of vegetation, willows and alders, as being quick growing and easily established trees, should be freely planted upon the banks. This fortunately is very easily done, for willow and alder sticks cut and put into the ground in the spring are pretty sure to do well. It is needless to say that the moister spots should be chosen for the willows, though they will do well in suitable soil in comparatively dry places. Besides giving shade and shelter to the fish, which is always an important consideration, a considerable quantity of food is bred upon trees and shrubs at the water side. I have found as many as eighteen caterpillars in the stomach of a trout which I caught under an overhanging oak tree.

CHAPTER IV

TROUT. PRELIMINARY HINTS AND ADVICE

The amateur who is beginning trout culture had better by all means buy eyed ova from a fish cultural establishment. There are many of these in the British Isles, and nowadays eyed ova are packed and sent safely all over the country. The artificial spawning of trout is not an undertaking in which the beginner is likely to achieve great success, and therefore I should advise him to avoid relying upon it when he commences his operations as a fish culturist.

Collecting the ova of wild trout is also an operation of some difficulty, and lays the beginner open to much more disappointment than if he deals with eyed ova purchased from a reliable establishment. Instead of having to watch and care for the ova through a critical and dangerous period, he receives them shortly before the young fish hatch out, when the ova are not in the most delicate stage.

It is of the greatest importance that everything should be ready for the ova long before they are expected, as hurry and new apparatus are likely to cause failure. Any concrete and varnished or enamelled woodwork should be exposed to the action of a current of water for at least five or six weeks before they are brought into actual use.

The choice of a suitable spot in which to make his hatchery is a serious point for the consideration of the amateur. A spring is the best water supply as a rule, for the water is usually of a fairly even temperature, and does not require filtering, but water from a stream where trout are known to live is quite safe. A few years ago it would have been necessary for any one wishing to take up fish culture, to erect a building in which to place his hatchery if he intended to hatch any number of eggs, in order to guard against frosts. At the present time, the eyed ova of even the brown trout (Salmo fario) can be obtained sufficiently late to be safe against a frost severe enough to cause any damage, and as the rainbow trout (Salmo irideus) spawns in February and March, the amateur is, at the time he receives the eyed ova, quite safe from frost.

The best method to pursue is to make long narrow ponds, with a current running through them, and to hatch the eggs out in trays and boxes suspended in these ponds. When the young fish hatch out, the trays which contained the ova can be removed, and the young fish kept in the boxes. Later on the young fish can be released from the boxes into the ponds. I shall subsequently describe how these ponds, trays, and boxes should be made.

The rearing ponds should be made, if possible, at a fall in the level of the water supply, so that they may be easily emptied. This is an important point which is frequently overlooked by amateurs. There should be an outlet on a level with the bottom of the pond, and if the water escapes through a pipe, that pipe should incline downwards. This, in a series of ponds, of course necessitates the ponds being at different levels, but the water is thus under much better control than if the outlet is at a higher level, and the ponds are easily emptied. Ponds may, however, be worked successfully with the outlet in mid-water, or even near the surface, though this does not ensure such a certainty of change of water throughout the pond. It is not, however, always possible to obtain such a difference in level between the supply and waste. In such cases the ponds should be made shallower near the outlet.

A popular idea seems to be that a gravel bottom is necessary for the well-being of trout; this is quite a mistake. Personally, I believe that a good earth bottom is best in a rearing pond, and even in a pond lined with concrete I should always put a layer of mould, preferably turf mould, at the bottom. With the use of this mould during the subsequent operations in rearing trout I shall deal later on.

The size of the ponds, of course, depends upon the number of trout to be reared. It is better to have several medium sized ponds than one large one, as then accident or disease occurring in a pond will only affect a portion of the stock of fish. Mr. J. J. Armistead in An Angler's Paradise, and How to Obtain It, says: "A pond sixty feet long, four feet wide, and about three feet deep, will hold ten or fifteen thousand fry at first, and give them plenty of room to grow, but by the end of July the number should be reduced to five thousand, which may be left till October, when they should again be