Spring.--This large and handsome duck has always been associated in my mind with the first signs of the breaking up of winter. Being a hardy species, it lingers on the southern border of ice and snow and is the very first of our waterfowl to start on its spring migration. We may confidently look for it in New England during the first warm days in February or as soon as the ice has begun to break up in our rivers and lakes. We are glad to greet these welcome harbingers of spring, for the sight of the handsome drakes flying along our water courses or circling high in the air over our frozen lakes, with their brilliant colors flashing in the winter sunshine, reminds us of the migratory hosts that are soon to follow. They are looking for open water in the rivers, for rifts in the ice or open borders around the shores of the lakes, where the first warm sunshine has tempted the earliest fish to seek the genial shallows, but they are often doomed to disappointment, for winter lingers in the lap of spring and again locks the lakes with solid ice driving the hardy pioneers back to winter quarters. The drakes are always the first to arrive and the females follow a few weeks later. Mr. Fred A. Shaw writes to me that in Maine--"The males generally make their appearance in March and in a short time select their mates, leaving early in April for their breeding grounds."

Courtship.--Perhaps the best account of courtship of this species is given by Dr. Charles W. Townsend (1916), as follows:

A group of five or six male mergansers may be seen swimming energetically back and forth by three or four passive females. Sometimes the drakes swim in a compact mass or in a file for six or seven yards or even farther, and then each turns abruptly and swims back. Again they swim in and out among each other, and every now and then one with swelling breast and slightly raised wings spurts ahead at great speed by himself or in the pursuit of a rival. The birds suggest swift motor boats by the waves which curl up on either side, and by the rapidity with which they turn and swash around. Again they suggest polo ponies, as one in rapid course pushes sidewise against a rival, in order to keep him away from the object of the quest. They frequently strike at each other with their bills, and I have seen two splendid drakes rise up in the water breast to breast, and, amid a great splashing, during which it was impossible to see details, fight like gamecocks. The pursuit is varied by sudden, momentary dives and much splashing of water.

The smooth iridescent green heads, the brilliant carmine bills tipped with black nails, the snowy white flanks and wing patches and the red feet, which flash out in the dive, make a wonderful color effect, contrasting well with the dark water and white ice. The smaller females, with their shaggy brown heads, their neat white throat bibs, their quaker blue-gray backs and modest wing patches, which are generally hidden, are fitting foils to their mates. I have reserved for the last the mention of the delicate salmon yellow tint of the lower breast and the belly of the male, a coloration of which he is deservedly proud, for, during courtship, he frequently raises himself up almost onto his tail with or without a flapping of the wings and reveals this color, in the same way that the eider displays his jet-black shield. Most of the time he keeps his tail cocked up and spread, so that it shows from behind a white center and blue border. Every now and then he points his head and closed bill up at an angle of 45 ^o or to the zenith. Again he bows or bobs his head nervously and often at the same time tilts up the front of his breast from which flashes out the salmon tint. From time to time he emits a quickly repeated purring note, "dorr-dorr" or "krr-krr."

The most surprising part of the performance is the spurt of water fully 3 or 4 feet long which every now and then is sent backward into the air by the powerful kick of the drake's foot. It is similar to the performance of the whistler but much greater, and while the foot of the whistler is easily seen and is plainly a part of the display, it is difficult to see the red foot of the merganser in the rush of water, although it is evident, doubtless, to the females. The display of the brilliantly colored foot in both species is probably the primary sexual display, and the splash, at first incidental and secondary, has now become of primary importance.

During all this time the female swims about unconcernedly, merely keeping out of the way of the ardent and belligerent males, although she sometimes joins in the dance and bobs in a mild way. At last she succumbs to the captivating display and submerges herself so that only a small part of her body with a bit of the crest appear above the water, and she swims slowly beside or after her mate, sometimes even touching him with her bill. Later she remains motionless, flattens herself still more, the crest disappears, and she sinks so that only a line like that made by a board floating on the water is seen. One would never imagine it to be a live duck. The drake slowly swims around her several times, twitches his head and neck, picks at the water, at his own feathers, and at her before he mounts and completely submerges her holding tightly with his bill to her neck meanwhile. Then she bathes herself, washes the water vigorously through her feathers and flaps her wings; the drake stretches himself and flaps his wings likewise. From the beginning of submergence by the female the process is the same in all the duck family that I have observed.

Nesting.--On the nesting habits of the American merganser there has been much discussion and many conflicting opinions, some asserting that it always nests in hollow trees or that it never nests on the ground. As a matter of fact it does both, for its nesting habits vary greatly in different localities or even with different individuals in the same locality. Major Bendire carried on an extensive correspondence with Mr. Manly Hardy on this subject and the latter was finally convinced that the goosander, as he called it, does occasionally nest on the ground. On June 12, 1891, Mr. Hardy wrote that, in passing through Caribou Lake, June 8, he found three nests, containing 26 eggs, on ledges under low fir bushes, which settled the controversy. I believe, however, that in the Eastern States and Provinces this merganser prefers to nest in hollow trees where it can find suitable cavities, which are usually scarce. Mr. Fred A. Shaw, who has had 30 years' experience with this species in the vicinity of Sebago Lake, Maine, contributes the following notes on its nesting habits:

A few breed around Sebago, especially near the mouth of Songo River, the principal tributary, where there is a large area of bog, flooded in spring, through which are scattered large hollow trees providing safe nesting places for them. The nest of this species is commonly placed in a hollow tree standing near the water and is composed of feathers and down from the breast of the parent bird. A nest of this bird near Whites Bridge at the outlet of Sebago Lake was shown me in May, 1897. It was in a white birch stub which was broken off about 15 feet from the ground and was hollow for about 10 feet from the top and contained 10 eggs, which were laid at the bottom of the hole on a warm bed of soft down from the breast of the mother bird.

Mr. Hardy wrote that it nested in hollow trees, usually hardwood trees, such as maples and ashes, and often in green trees. Mr. John H. Sage (1881) found a nest on an island in Moosehead Lake, Maine, on June 19, 1881--

in a hollow under the roots of a standing tree, roots, earth, and moss forming a perfect roof, so that the nest, after the heavy shower of that day, seemed well protected and was quite dry. The eggs were covered with leaves, moss, and feathers--mostly feathers. The old bird was seen to leave the nest.

On Lake Winnipegosis and Waterhen Lake, Manitoba, we found the American merganser very common and nesting on the numerous islands, wherever suitable nesting sites could be found among the piles of loose boulders along the shores. Mr. Walter Raine found as many as 30 nests of this species on Gun Island, a large island in Lake Winnipegosis. On one island that we visited the Indians had collected about 60 eggs of the "saw bills," as they call them, a short time previously. The nests were very well hidden in remote crevices among the piles of large boulders; many of them were quite inaccessible, where the boulders were too large for us to move them. We often saw mergansers flying away from islands where we felt sure that they were nesting, but where we were unable to find or reach the nests. A few nests were found in the dense tangles of gooseberry bushes and nettles on the tops of the islands where hunting for them was difficult and painful unless a telltale path, strewn with feathers and droppings, told us just where to look.

Near the north end of Lake Winnipegosis, on Whiskey Jack Island, we visited, on June 18, 1913, a deserted ice house where we were told that we might find the "little saw bills," hooded mergansers, nesting. It was an old tumbled-down affair, with the roof nearly gone and partly filled with loosely piled bales of hay; there did not seem to be any suitable nesting sites for "saw bills" anywhere in the vicinity, so we sat down to eat our luncheon. While so occupied we were surprised to see a female American merganser fly up and alight on the landing and gaze longingly into the ice house. We then began an exhaustive search by moving the bales of hay and crawling into the crevices between them. While peering into a dark cavity I though I saw something moving regularly like a breathing duck; we pulled away some more bales and there sat a female merganser on her nest within 2 feet of my face; I reached in to catch her but she slipped away and escaped through another passage way. There were 15 eggs under her in a nest profusely lined with white down, mixed with hay. Further search revealed another nest near by, similarly located; the bird had left the nest and had carefully covered the 12 eggs which it contained with a soft blanket of down. Lieut. I. T. Van Kammen (1915) found two nests in an old, abandoned lighthouse tower; the nests were about 3 feet apart and "each nest was placed in a depression, perhaps 5 inches deep, scraped out of the soft dirt of the lighthouse floor."

Audubon (1840) gives an attractive account of finding the nest of a goosander on a marshy island. He describes the nesting site and nest as follows:

The islands on which the goosander is wont to breed are mostly small, as if selected for the purpose of allowing the sitting bird to get soon to the water in case of danger. The nest is very large, at times raised 7 or 8 inches on the top of a bed of all the dead weeds which the bird can gather in the neighborhood. Properly speaking, the real nest, however, is not larger than that of the dusky duck, and is rather neatly formed externally of fibrous roots and lined round the edges with the down of the bird. The interior is about 7 1/2 inches in diameter and 4 inches in depth.

Mr. W. L. Dawson (1909) says of its nesting habits in Washington:

Now and then a crevice in the face of a cliff does duty, and old nests of hawk or crow have been pressed into service. Moderate elevations are favored, but Mr. Bowles once found a nest near Puget Sound in a decayed fir stub at a height of over a hundred feet. The cavity, wherever found, is warmly lined with weeds, grasses, and rootlets, and plentifully supplied with down from the bird's breast.

Mr. Fred H. Andrus (1896) thus describes a nest which he found in Oregon:

May 26, 1895, I collected a set of 10 American merganser's eggs from a hole in the rocks about 100 feet above the Umpqua River. The nest was about 15 feet from the top of a nearly perpendicular cliff about 50 feet in height, and was found by watching the bird. In going to the nest the bird would fly up and down the river in an oval course several times, and finally, coming close to the water as if to light, would rise to the nest. The entrance to the hole was 6 inches by 12, and the inside dimensions 4 feet long, 2 feet deep, and 18 inches high. The nest was about 1 foot in diameter, of down mixed with moss, one-half inch thick in the center and thicker around the edges.

Eggs.--The American merganser raises but one brood in a season and lays from 6 to 17 eggs; some writers say it lays from 6 to 10 eggs, but I think these small sets must have been incomplete or second attempts; I should say that the commonest numbers would run between 9 and 12, but I have personally taken sets of 15 and 16. The eggs are usually distinctive, and typical eggs are not easily mistaken for anything else. The female is difficult to distinguish from the female red-breasted merganser when she flies from the nest, though she is a decidedly heavier looking bird and has more white in the wings. The down, however, in the American merganser's nest is much whiter than that in the nest of the red-breasted and the color of the eggs is different. The down is grayish white in color, about No. 10 gray of Ridgway, or "pale gull gray," and it is usually mixed with numerous pure white breast feathers and considerable rubbish or bits of straw. The eggs of the American merganser are very pale buff or "ivory yellow." The shape varies from elliptical ovate to elliptical oval.

The measurements of 93 eggs in various collections average 64.3 by 44.9 millimeters; the eggs showing the four extremes measure 72 by 46, 64 by 50, 55.4 by 38.5 and 56.7 by 37 millimeters.

The drakes desert the ducks and usually disappear from the breeding grounds entirely as soon as the eggs are laid, leaving the females to perform the duties of incubation and care for the young alone. In Newfoundland we saw only females on the lakes, where they were busy with family cares, but we saw plenty of males on the swift water rivers, playing in the rapids and fishing in the pools. Several observers in Maine have said that the males are not seen during the summer, but this may be due to the fact that the males are in eclipse plumage at this time and are very shy and retiring. Mergansers which nest in hollow trees are usually very close sitters, and it is often impossible to drive a sitting bird from her nest by pounding the tree; on the other hand, those which nest on the ground on islands usually slip away long before the intruder reaches the vicinity of the nest, often before a boat lands on the island; a deviation from either of these habits would, in either case, tend to reveal the nest. Fresh eggs, taken from incomplete sets in Manitoba and hatched in our incubators, showed an incubation period of 28 days.

Young.--Several writers state that the young mergansers are carried from a nest in a hollow tree to the water, in the bill of the parent bird; Millais (1913) says that Mr. Oswin Lee has seen a female goosander carry down nine young ones out of the nest, and that she "carried them partly in her beak, partly between the beak and the breast."

Mr. Shaw, however, offers the following evidence to the contrary:

An interesting occurrence in connection with the breeding of this bird was related to me by Mr. G. H. Moses, who while in camp at Songo River had exceptional opportunity to observe them. In the spring of 1896 a nest was located in a tall hollow tree where it could be readily seen from his camp door. After the young were hatched Mr. Moses saw the mother bird alight in the water at the foot of the stub in which the nest was located and commence to call to the young birds in the nest. Immediately the little ones came tumbling down one after another from the hole in the tree top to the water and at once swam away with their mother.

Mr. William S. Post (1914) has twice witnessed a similar performance; the following is his account of it:

It was my good fortune to witness twice the emerging of a young brood of mergansers from an extreme situation of this kind, an old pileated woodpecker's hole about 40 feet high in the limb of a live elm, standing about 15 feet from the edge of the Tobique River in New Brunswick.

One June 18, 1910, I fished the famous salmon pool at the fork of the river, and having incidentally run the canoe close to the shore near where this old elm stands, I landed and rapped several times sharply on the tree with a stick, for I had been told that a wood duck--which on the Tobique means a golden eye--nested there the previous spring. The female merganser immediately flew out and, having circled about over the river, alighted on the water. After assuring myself of the identification, which caused me some astonishment on account of the size of the bird in proportion to the entrance of the hole, I returned to my fishing.

In a few moments I noticed a small bird drop down apparently from the hole, and in a few more seconds another and then a third. My first thought was that a bank swallow, of which there are many on the river, had flown up near to the hole and down again three times in succession. This caused me to stop fishing to watch, when to my astonishment a small bird with white breast appeared in the hole, jumped out, and was followed by another, and again another. I then lost no time in reaching a point in the river opposite the tree, where I saw in the water against the bank, swimming around, a brood of 11 young ducks. I was much surprised, as I had been under the impression from what I had read that the old duck would certainly carry down the young from such an inaccessible position, and though I believe the young birds must have landed in the water, I was yet astonished that they could withstand the shock of such a drop, and I presumed that by rapping on the tree I had caused the old bird to leave in such fright that her fear had been communicated to the young and they had followed her example, and that the whole procedure was therefore an unnatural one.

The clubhouse is situated directly across the river, and on June 12, 1913, two years later, I was sitting on the piazza when my attention was attracted by seeing something large drop from the top of this same elm into the water. I immediately saw that it was the old sheldrake and that she was swimming around close to the shore.

In a few seconds another dropped from the hole to the ground and I could see it run down the bank and join its mother who was calling loudly and turning round and round in the water. This one was quickly followed by others in succession until there were seven. By this time I had called my guide and in company of one of the members of the club was crossing the river, provided with trout landing nets.

The old bird seeing us immediately swam upstream and around the point with her brood and this was the last we saw of her. We landed and stood under the tree where we could hear distinctly more young ducks peeping in the hole. Looking up we saw one tottering on the edge, and before we could take stations where we could properly observe the actual drop he had struck the ground close to my friend and made such rapid progress toward the water that he escaped in spite of landing nets. In a few seconds another, which proved to be the last, followed, falling on the other side of the tree, and I promptly made him a captive. The first bird was in the water and had immediately dived. It is strange that he should have known enough to seek the water, and also to dive immediately.

After a day or two of rest in the nest, probably longer in tree nests than in those on the ground, the young have dried off their down and gained sufficient strength to take to the water, where they are very precocious. The downy young are very handsome and attractive. It is a beautiful sight to see a female merganser swimming in the clear calm water of some mountain lake or wilderness stream, where the mirrored reflections of picturesque scenery and forest trees make a splendid setting for the picture of a swiftly gliding, graceful duck followed by a procession of pretty little balls of down, with perhaps one or two of them riding on her back. If danger threatens she quickens her pace, but the little fellows are good swimmers and keep right at her heels; even if she dives they can follow her under water, working their little paddles vigorously and darting along like so many fish. If too hard pressed she rises and flaps along the surface, half flying; they can almost keep up with her at this pace, for they can run along the surface as fast as we can paddle our canoe. They soon become exhausted with some exertion, so she leads them into some sheltered cove, where they can run up on the shore and hide in the grass, or even up into the thick woods, where it is almost hopeless to hunt for them; it is surprising to see how quickly the young and even the mother bird can disappear. Millais (1913) relates the following interesting incident:

When rushing down the swift rivers of Newfoundland in my canoe, I have often wondered at the resource of natural instinct of the broods of goosander and their mothers which remain perfectly still when suddenly confronted with danger. As the little boat flies down a rapid, swiftly passing silent pools in the rock eddies at the sides, I have often turned my head and noticed a female goosander and her nearly full-grown young. On a lake or open stretch of the river, knowing that concealment was impossible, the mother would have dashed out in the open, and either hurried by flapping along the surface to the middle of the lake, or, in the case of the river, downstream, and so endeavor to escape. When suddenly confronted within a few yards in the eddies of the rapids, she felt that such a method of escape was useless, and with swift intuition remained perfectly still, each member of the brood keeping the neck held stiffly, so that the whole party looked like the stiff twigs of an upturned tree. This sudden assimilation to surroundings, so wonderfully exhibited in the common or little bittern amid the rushes, seems to be a natural instinct in all birds, and they often adopt it as a last resort.

Plumages.--Downy young mergansers are beautiful creatures; the upper parts, including the crown, down to the lores and eyes, hind neck, and back, are rich deep "bister" or "warm sepia," relieved by the white edging of the wing and a large white spot on each side of the rump; the sides of the head and neck are "mikado brown" or "pecan brown," shading off on the neck to "light vinaceous cinnamon" or "buff pink"; a pure white stripe extends from the lores to a point below the eyes and it is bordered above and below by dark brown stripes; the rest of the lower parts are pure white. The nostril is in the central third of the bill, instead of in the basal third, as in the red-breasted merganser.

In juvenal plumage, which is acquired at about the time that the young bird attains its growth, the sexes are practically indistinguishable except that the male is slightly larger. This plumage is similar to that of the adult female except that, in the young bird, the white throat extends down to the chest, whereas, in the adult female, the lower throat is brown. The wing pattern is also different in young males, which have the outer secondaries white and the inner secondaries gray. During the fall and winter an almost continual molt is in progress, black feathers appearing in the head and neck, producing a mottled effect, and vermiculated feathers appearing in the flanks. The tail is renewed in the spring but not the wings. The first postnuptial molt, which can hardly be said to involve an eclipse plumage, takes place in August and September; this molt is complete and is prolonged through October at least; by November, when the bird is nearly a year and a half old, the adult plumage is complete.

Millais (1913) says:

In May the adult male goosander begins to assume its eclipse plumage. The adult male in August has the crown reddish brown, with a gray tinge; chin white, and the rest of the head and upper neck rich red brown. There is a black mark in front of the eye, and a whitish line from this to the lower angle of the upper mandible. The lower neck is blue gray, interspersed with creamy white; mantle, flanks, scapulars, back, and tail blue gray; the flanks have a few white feathers on the outer sides, vermiculated with brownish gray; the last inner secondaries only change from black to black; wings as in winter, and now changing as usual only once; under parts not so rosy as in winter. In early September the wings and tail are renewed, and the black feathers of the mantle come in. After this the whole plumage proceeds to molt slowly, the full winter dress not being assumed until early December.

Subsequent molts of adult males consist of a postnuptial molt of the contour feathers early in the summer into the eclipse plumage, a molt of the flight feathers in August or September, and a complete molt of the contour feathers out of the eclipse plumage in the fall. Females probably do not make the double molt of the contour feathers but have a complete molt in the late summer.

Food.--The merganser is primarily a fishing duck, at which it is very skillful and a voracious feeder. It pursues under water and catches successfully the swiftest fish. Often a party of sheldrakes may be seen fishing together, driving the panic-stricken fish into the shallows or into some small pool where they may be more easily caught. Mr. Hardy writes:

They fish in companies; as fast as they come up the hind ones run ahead of those in front of them and dive again, being in turn succeeded by others. I have seen them fishing on quick water in very cold weather until January 7. They feed exclusively on fish, several often uniting to capture one of large size. Last year we took a pickerel from a party of them which measured 14 inches in length; also took from one's throat a chub which, with head decomposed, measured 10 inches.

One of these gluttonous birds will often attempt to swallow a larger fish than it can dispose of, leaving the tail of the fish protruding from its mouth while the head is digesting. Mr. Shaw found in the "stomach" of one bird "13 perch, a few of which were nearly 3 inches in length."

Audubon (1840) says:

I have found fishes in its stomach 7 inches in length, and of smaller kinds so many as to weigh more than half a pound. Digestion takes place with great rapidity insomuch that some which have fed in captivity devoured more than two dozen fishes about 4 inches in length, four times daily, and yet always seemed to be desirous of more.

Mr. Harry S. Swarth (1911) describes the following manner of feeding, which is unusual for this species:

I was concealed in the shrubbery at the water's edge examining a large flock of ducks for possible rarities, when a dozen or more mergansers (both M. americanus and M. serrator) began swimming back and forth but a very short distance from my blind. They swam slowly, with neck outstretched, and with the bill held just at the surface of the water, and at a slight angle, so that the head was submerged about to the level of the eyes. The water was evidently filtered through the bill, as a slight "gabbling" noise was quite audible, and obviously something was being retained as food, though just what it was I could not tell. This is rather remarkable, as it is exactly the manner of feeding usually employed by the shoveller (Spatula clypeata), a species which, as regards bill structure, is further removed from the mergansers than any other member of the Anatidae.

Mr. Ora W. Knight (1908) says:

Along the coast in winter they eat many mussels and allied species of mollusks, swallowing them shell and all. The shells are soon ground to pieces in their intestines and stomachs, and in dead birds dissected out I have traced the entire process from entire mussel shells down to impalpable mud at the lower end of the intestinal tract.

In the early spring, when live fish are difficult to obtain, they seem to enjoy frozen, rotten fish with the same gusto as fresh, picking them out of the floating ice. They also feed to some extent on frogs, small eels, aquatic salamanders, crawfish, and other small crustaceans, various bivalve mollusks and snails, leeches, worms, water insects and larvae, and the stems and roots of aquatic plants.

Behavior.--The American merganser is a heavy-bodied bird and sometimes experiences considerable difficulty in rising from the water; if the circumstances are not favorable, it has to patter along the surface for a considerable distance; when flying off an island it often does the same thing unless it gets a good start from some high place, so that it can swoop downward. In swift water it has to rise downstream, as it can make no headway against the current; but it generally prefers to fly upstream if it can. Mr. Aretas A. Saunders writes to me, in regard to the flight of mated pairs, noted in Montana, "that they flew off, with the male in the lead in each case," also "that they left the water flying in a long, low slant upstream, not rising high enough to see them above the willows that lined the stream until they had flown a considerable distance." When well under way the flight of this species is strong, swift, and direct; on its breeding grounds it usually flies low, along the courses of rivers or about the shores of lakes, seldom rising above the tree tops; but on its migrations it flies in small flocks, high in the air with great velocity. The drake may be easily recognized in flight by its large size, loon-like shape, its black and white appearance above, dark green head and white under parts; its flight is said to resemble that of the mallard. The female closely resembles the female red-breasted merganser, but it is a more heavily built bird, has a more continuous white patch in the wings, the white tips of the greater coverts overlapping the black bases of the secondaries, giving the appearance of a large white speculum, whereas in the red-breasted merganser the black bases of the secondaries show below the greater coverts, forming a black stripe through the middle of the white speculum. When flying to its nest cavity in a tree or cliff it rises in a long upward curve and enters the hole with speed and precision. Mr. Harry S. Swarth (1911) refers to--

a peculiar habit which made this species quite conspicuous throughout the summer, was that of individuals rising high in the air and circling about for hours at a time, uttering at frequent and regular intervals a most unmelodious squawk. Both sexes were observed doing this, and the habit was kept up until about the end of August.

This sheldrake is probably the most expert diver of its tribe, being built somewhat like a loon and approaching it in aquatic ability. It can sink quietly down into the water like a grebe or dive quickly with a forward curving plunge, clearing the water for a foot or more, as it does so. It swims swiftly on the surface, but can attain even higher speed below it, where few fish can escape it. Dr. Charles W. Townsend (1909) infers from its method of diving, that the wings are not used when swimming under water and he quotes a statement from Selous to the same effect; he says:

The American and the red-breasted merganser both dive like the cormorant. They often leap clear of the water, in graceful curves, with their wings cleaving closely to the sides. At other times the leap is much curtailed, or they sink beneath the surface without apparent effort. I should infer, therefore, that the wings were not used under water.

On the other hand I quote from Mr. Walter H. Rich (1907) as follows:

Seen under the water in pursuit of a breakfast or dodging about to escape capture when wounded, the resemblance to some finny dweller of the sea is very marked--head and neck outstretched, every feather hugged closely to the body, the half-opened wings like large fins aiding the feet in their work, he goes shooting through the water like a flash.

Probably both observers are correct, for birds are not bound by hard and fast rules. The rapidity with which this species can dive from the air is remarkable. While in full flight it plunges into the water, swims below the surface for a distance and then suddenly emerges and continues its flight. Millais (1913) says of its behavior on land:

The walk is very heavy and rolling, and the feet are placed on the ground deliberately, whilst the bill is pointed downward, and each step taken is as if the bird was afraid of tripping or falling. They seldom go more than a yard or two from the water's edge, but can run quite swiftly for a few yards if suddenly surprised. In winter it is a very rare event to see goosanders ashore, but in spring they often leave the water, and will spend hours sleeping and preening on some small island or point of land. No birds are more industrious in their toilet than the mergansers in spring, and most of their time, when not feeding, flying, or sleeping, is spent in polishing up their plumage and bathing.

Of the vocal performances of this species I know very little; I have never heard a sheldrake utter a sound, so far as I can remember, and very little seems to have been written on the subject. Audubon (1840) describes the notes of the goosander as "harsh, consisting of hoarse croaks, seldom uttered unless the bird be suddenly startled or when courting."

Game.--As they live almost exclusively on fish, sheldrakes are not considered good table birds and so are not much sought after by gunners. But young sheldrakes are not unpalatable, and many gunners shoot them regularly for food. They do not come to decoys, but, as they are swift fliers and hard to kill, shooting them is good sport. Many sportsmen feel justified in killing them on account of the large numbers of trout which they consume; but this is hardly justifiable, for they also destroy many predatory fish, such as pickerel and thus help to preserve the balance of nature.

Fall.--The name "pond sheldrake" has been applied to this species because it shows more preference for fresh water than its relative, the red-breasted merganser; its fall migration is more inland, where it flies along our larger water courses and frequents our lakes and ponds until it is forced coastwise by the freezing of its favorite resorts.

Winter.--But even in winter it still lingers wherever it can find open water near its summer home and its migration is one of the shortest. Mr. Shaw says that, at Sebago Lake, Maine--

a few remain through the winter in the coldest weather spending the day in the open water at the foot of the lake and in the upper part of Presumpscot River, its outlet, and at night leaving for the salt water.

It is more common on the coast of New England than in the interior in the winter, but it winters in large numbers in some of the Great Lakes and on the large rivers of the interior, especially in the rapids and about the cascades of clear-water streams. The icy waters of our northern streams have no terrors for this hardy fisherman provided it can find open water and plenty of food. Rev. Manley B. Townsend writes to me:

Every winter during my residence in Nashua, N. H., 1912-1918, I noted considerable numbers of these "fish ducks" as they are locally called, between Manchester and Nashua. Even in the coldest, most inclement weather, when the Merrimack was frozen several feet deep, these birds could be seen sitting on the ice about the rapids, where the swiftly flowing water kept the river free from ice, or swimming and diving in careless abandon. I once counted 40 on the ice or in the water at a single rapid. Fish seemed to be abundant. The birds apparently wintered well.