On any list of North American birds selected for their general familiarity, the song sparrow would have few peers. Although relatively small and not very conspicuously marked, this species combines a rediness to dwell near humans and a persistent and attractive song with a breeding range extending from Mexico to the outer Aleutians and from the islands off the Atlantic coast to those off the coast of the Pacific. Further, the territories of the males are small and the suitable habitats extensive, with the result that the song sparrow is abundant in most of its range. Add to the foregoing the fact that these sparrows are readily trapped, and it is not surprising that some populations have been studied in meticulous detail.

Not only is the song sparrow one of our best-known birds; it is also our most variable, with 31 subspecies recognized as occuring within the territory covered by the A.O.U. Check-List (1957) and 3 additional subspecies in Mexico (Friedmann, et al., 1957).***

The breeding song sparrow of eastern Canada and of the United States west to the Appalachians displays the typical preference of this species for moist ground and for a low, irregular, dense plant configuration considerably exposed to the sun. "No land bird seems more fond of water," writes E. H. Forbush (1929). Everywhere it is "primarily a bird of the lower lands. . ." (Knight, 1908), along the banks of streams, the brushy shores of ponds, and in the shrubby wet meadows or cattail swamps. Even on the central Atlantic coast, where the race atlantica replaces the present subspecies on the beaches, melodia seems to be the song sparrow found back in the thickets (Stone, 1937), and Dexter (1944) has reported a nest of this race in a salt marsh on a tidal inlet at Gloucester, Mass. But these lowland situations are only a first preference, for the bird is tolerant of a wide range of conditions. It is often found in brushy fence rows and along country roads; and it sometimes breeds even in rocky wooded clearings in Maine, in small wooded openings only a few rods in diameter in New York (Eaton, 1914), and in second-growth woodland in Pennsylvania (Todd, 1940). Gardens and yards offer the song sparrow sunny, bushy, moist cover, and the bird is a common nesting species in suburbs and small towns. On Mount Mitchell, North Carolina, the highest point in the eastern half of the United States, Burleigh (1941) observed individuals of the race euphonia as high as 6,300 feet above sea level during the summer.

Spring.--Although a few song sparrows winter far north, most withdraw from that part of the range in Canada and northern New England. M. M. Nice (1933) has cited the evidence, from banding records, that some individuals of this subspecies are resident "in regions where most of their kind are migratory." Hervey Brackbill (1953) has found that the breeding population of Maryland contains both migratory and sedentary individuals. In the Hudson River valley and around New York City the species is a common winter resident. The arrival of migrants, and the return to their breeding territories of birds that have wintered, normally begins in the latter half of February in the southern part of the range, while the first spring arrivals appear in Maine in mid- and late March and in Canada in March and early April. In Maine migration continues into early May (Palmer, 1949).

Carl H. Helms (1959) weighed song sparrows captured in Massachusetts just before and just after a large night migration in April. Birds weighed before the flight averaged 1.41 grams heavier than those that arrived as a result of the movement. These post-flight individuals were noticeably less fat. The song sparrow sings even on cold clear mornings of late winter, and its voice is a characteristic sound of early spring. Aretas A. Saunders (1947) reports singing in Connecticut when the temperature was -2^o F.

Territory.--Territory appears to be established principally or entirely by the male. E. H. Forbush (1929) has described behavior that probably includes elements both of territorial defense and of courtship or pair formation: "There is considerable rivalry among the males, but their contests appear to be mainly competitions in song and flight. They chase the females and each other about through the air with fluttering wings, often sailing and singing. Their pursuit seems not to be in earnest, as, notwithstanding the rapid movement of their wings, their progress is slow. Now and then a bird pauses in his flight to sing, supported for an instant on his widespread pinions. Flight-songs also carry them up into the air. Occasionally a battle ensues between two rival males, and sometimes they even roll and tumble in the dust with locked bills and beating wings." In the section on courtship is a description of behavior that probably has territorial functions as well. The song sparrow's persistent songs, six to eight per minute at dawn in spring (Forbush, 1929), are, of course, associated with the maintenance of territory.

The size of the individual territory in favorable habitat is less than an acre. Robert E. Stewart and Chandler S. Robbins (1958) give an interesting series of data on the population density of breeding song sparrows in Maryland. (Population densities provide a basis for estimating only maximum territory size, for it does not follow that all the area censused actually fell within the boundaries claimed by the males.) In 19.2 acres of " 'shrubby field with stream-bordered trees' " were 21 territorial males; in 9.5 acres of " 'open hemlock-spruce bog' (brush-meadow stage. . .)" were 3 males; and in 20.5 acres of " 'moderately sprayed apple orchard with infrequently mowed ground cover. . .' ") were 4.5 territorial males.

Courtship.--Witmer Stone (1937) writes as follows of his observations on Cape May: "In late March and April the air seems simply filled with Song Sparrow song and at this time we see male birds flying from bush to bush with neck stretched out, head and tail held high, and wings vibrating rapidly. This seems to be a part of the courtship display and as soon as the bird alights it bursts into song. On March 21, 1925, and April 2, 1914, I have noted this performance and the birds were evidently paired. . . ." This behavior probably was associated also with territory defense. To the account from Forbush quoted in connection with territory may be added the same author's statement that song sparrows "spend much time in the pleasant pastime of courtship. The females seem to be modest and coy." The duration of the periods of pair formation and between pair formation and nesting seem not to have been recorded. A comparison of the dates given for the height of the return of spring migrants and for the beginning of general nesting in a given locality suggests that a month or more often elapses in these "prenuptial" and "preliminary" periods (Nice, 1943).

The same males and females have been found mated to each other in successive years (Hamill, 1926; Higgins, 1926).

Nesting.--Building is carried on principally by the female, but Ora W. Knight (1908) once saw a male apparently assisting his mate. He was "more inclined to shirk his share, picking up material, dropping it and picking it up again, singing meanwhile." During building, says Forbush (1929), "the male devotes himself more to song than to labor."

The duration of building is variously given, with 5 days the lowest figure and 10 days a commonly accepted maximum. Weather is known to influence the speed of building; and it may be supposed that the time-advance of the season, the number of nesting attempts already made, and the presence or absence of an earlier brood might all affect the female's building behavior. As mentioned below, females of the race melodia commonly raise three broods in a season and sometimes deposit the eggs for a later brood in a previously successful nest. Andrew J. Berger (1951) found five nests built in one season by the same female of the subspecies euphonia; not all succeeded.

The heights of nests range from ground level to at least 12 feet high. Most nests are placed on the ground, usually concealed under a tuft of grass, a bush, or a brush pile; and elevated structures are rare or absent early in the season. Eaton (1914) reports that 99 percent of the nests in New York are on the ground, but most writers use words suggesting only that something over half are located there. Elevated nests are "at a height of generally not over two or three feet" (Knight, 1908), but numerous references to somewhat higher sites can be found. Locations over water are not uncommon.

Plants in which nests are placed are grasses, sedges, cattails, a great assortment of bushes and shrubs, and, more rarely, trees of many species. Nests are occasionally built in cavities. Hollows in old apple trees are apparently the commonest such locations (Knight, 1908; Todd, 1940; Eaton, 1914); but hollow logs and rails, unoccupied buildings such as woodsheds, and even nest boxes (Palmer, 1949) have been resorted to. The materials used for the outer, bulky part of the nest, as opposed to the lining, are most commonly leaves, strips of plant bark, and weed and grass stems. The lining is of fine grasses, rootlets, and horse or other animal hair. W. E. C. Todd (1940) states that nests "when above the ground. . .are often quite bulky." Knight reports the dimensions of one nest placed on the ground: the diameter of the cavity was 2 1/2 inches, while the overall diameter varied between 5 and 9 inches; the cavity depth was 1 1/2 inches, the overall depth 4 1/2 inches.

Eggs.--The statements in this paragraph are applicable to the song sparrow without regard to race. The female lays from 3 to 6 eggs. They are slightly glossy and range from ovate to short ovate. The ground color of freshly laid eggs is "pale Niagara green" but this fades upon exposure to a greenish-white. Most eggs are very heavily speckled, spotted, or blotched with reddish browns such as "Verona brown," "russet," "cinnamon brown," or "Brussels brown." Some eggs have undermarkings of "pale neutral grey." The spottings generally are more or less evenly distributed over the entire surface, sometimes obscuring the ground color and making it appear to be a light buffy brown. They vary considerably both in shape, size, and intensity. The measurements of 400 eggs average 20.4 by 15.6 millimeters; the eggs showing the four extremes measure 25.9 by 17.8, 24.5 by 18.3, 16.9 by 15.4, and 19.6 by 12.8 millimeters.

Turning to the race melodia, the measurements of 50 eggs average 19.5 by 15.1 millimeters; the eggs showing the four extremes measure 21.9 by 14.5, 20.1 by 16.8, 17.8 by 14.2, and 18.0 by 14.0 millimeters.

Incubation.--Forbush (1929) states that incubation is "by both sexes, female chiefly" and that "some males assist the females a little. . . ." However, there is no evidence that males have an incubation patch, without which they would be ill-equipped to supply heat to the eggs. Mrs. Nice (1937) found males of the race euphonia lack this patch, and her unequivocal statement that only the female euphonia incubates casts doubt on Forbush's contention. The role of the male during incubation is probably confined to the defense of territory and nest.

Incubation seems to start sometime not many hours from the laying of the last egg of the set, if this inference may be drawn from the failure of observers to report differences in development in the young of a brood. In euphonia Mrs. Nice observed that most often a clutch hatched over a 2-day span, indicating that incubation had begun before all eggs were laid.

The incubation period is said by Knight and Forbush to be from 10 to 14 days, but it is doubtful if accurate measurements of the period, as it is presently defined, would be less than 12 days, as in euphonia (Nice, 1937). W. E. Schantz (1937) watched a female euphonia incubate three eggs (laid July 10 - 12) for 24 days; the eggs failed to hatch.

Young.--Most of our knowledge of the development of the behavior of nestling song sparrows comes from Mrs. Nice's work, devoted chiefly to euphonia. The following paragraph is based on her report (1943). The development of the plumage is described below under the heading Plumage.

Newly hatched song sparrows can grasp, gape, swallow, defecate, and change location "by means of uncoordinated wrigglings." A feeding note has been heard in 2-day-old birds. The eyes begin to open at age 3 or 4 days. Incipient preening motions appear at age 5 days, as do, rarely, cowering and the ability to utter a location call. At age 7 days many motor coordinations are acquired, and henceforth the bird "is capable of leaving the nest." Among the behaviorisms of the 7-day-old are cowering, stretching of the wings, head-scratching, yawning, and climbing to the nest rim. Birds 8 and 9 days old acquire new types of wing-stretching, engage in wing-fluttering and -fanning, and body-shaking, and utter new feeding notes.

Both parents feed the nestlings, chiefly on "insects, worms, beetles, grubs, flies, caterpillars, grasshoppers, and similar insects" (Knight, 1908). The period in the nest varies, its minimal limit being given as 7 days by Forbush (1929) and its maximum as 14 days by most writers. Seven days undoubtedly does not represent a natural, undisturbed nestling period, but is probably the youngest age at which nestlings will leave the nest when disturbed. Knight says that young leave ground nests earlier than they do elevated nests, and that this early age is 10 days. At this time they are still unable to fly, and newly fledged birds remain hidden in plant cover. Mrs. Nice (1937) states that young euphonia "when. . .about 17 days old. . .are able to fly and come out of hiding."

Dependence on the parents continues until after the post-juvenal molt (Todd, 1940). The parental bond may be assumed to be severed at the age of about 28 to 30 days as in euphonia (Nice, 1937).

As in other species, juvenile song sparrows occasionally engage in some of the behavior of nest building (Hoyt, 1961).

As second and third broods are produced regularly, and fourth broods probably occasionally, as far north as Massachusetts, the matter of timing successive families is of interest. "When the young of the first brood are able to fly, the female immediately begins to deposit eggs for the second brood, often in the same nest, leaving the male to care for the first, and he attends them usually until the young of the second brood have hatched, when he leaves them to help feed and care for the younger brood (Forbush, 1929)."

Plumages.--In the following description of the plumages and molts, material involving both melodia and euphonia has been used. The sexes are identical in their molts and practically identical in their plumages. Females average a little later than males in date of molt. Minor sexual differences in plumage will be mentioned.

Natal down, described as both sepia-brown in color (Dwight, 1900) and black (Nice, 1943) is present at hatching. Mrs. Nice (1943) writes that this down "is prominent on the dorsal, femoral and occipital regions and on the coverts. For the first two days there is little change except in increased length of down." The progress of the molt into the juvenal plumage is described in detail by G. M. Sutton (1935), who writes:

. . .the nestling-stage of the juvenal plumage is. . . notable for its dullness, the feathers of the loral, malar, and superciliary regions being still for the most part in their sheaths, and the tertials so short that their rich edgings are not yet apparent. The streaking of the chest is quite sharp, but on the sides it is, if anything, less marked than in later stages. Male and female birds are apparently not distinguishable at this age. The pectoral streaking is so much intensified because the feathers lie close together and are partially sheathed, that the actual width of the streaks is difficult to determine.

By the time the tail is an inch long the feathers of the face are almost altogether unsheathed, the tertials and secondaries are practically of full length, the pectoral plumage is fully fluffed out, and the bird is, therefore, much more colorful in appearance. At this stage males may, with a fair degree of certainty, be distinguished from females by the heavy streaking of the chest. Chapman. . . tells us that the "breast blotch is wanting" in this plumage. While this is no doubt to a considerable extent true, two individuals in a series of eleven specimens at hand show a definite blotch and two others exhibit a tendency toward convergence of streaks in the middle of the chest.

Sutton considers that song sparrows have a rather definite and complete juvenal plumage. "By the time the juvenal rectrices are of full length the body plumage is comparatively complete with all the feathers unsheathed and with no noticeable intrusion of pin-feathers of some subsequent plumage." He says that "specimens in juvenal feather may be taken during a long period of the summer," but is cautious about concluding how long the individual bird wears the plumage.

The foregoing is essentially an account of the molt, not the plumage, of which Dwight (1900) gives a good description. The bird "resembles Z. albicollis, but lacks chestnut above" and is "paler on [the]crown and less streaked below. Above, including sides of head, wood-brown or sepia broadly striped on back, narrowly on crown, nape and rump with dull black, the feathers centrally black with a narrow zone of walnut and wood-brown and grayish edgings. Indistinct median crown and superciliary stripes dull olive-gray with dusky shaft streaks. Rictal and submalar streaks black; orbital ring buff. Wings dull black with walnut edgings, the wing coverts and tertiaries buff tipped. Tail olive- brown broadly edged with walnut and indistinctly barred. Below, dull white washed with pale or yellowish buff deepest on the throat and flanks and streaked on sides of chin, throat, breast and sides with dull black. Feet and bill pinkish flesh, becoming dusky with age, lower mandible remaining partly flesh-color." Dwight believes this plumage is worn several months; it fades considerably.

The first winter plumage is acquired, according to Dwight, "by a partial, sometimes complete, postjuvenal moult" beginning in some birds in mid-August, in others not until the last of September. These latter will still show new feather growth late in October or early in November, although "the whole period of moult does not cover much more than two months in the majority of cases." The molt "involves the body plumage and the tail and very often, part at least, of the remiges. The renewal of five or six outer primaries occurs in nearly all young birds of this species and is very likely characteristic of the first brood. . . . The secondaries are rarely found in moult, the tertiaries, alulae and wing coverts regularly so. . . . [Occasionally] the renewal of primaries, secondaries and even of rectrices, might easily be overlooked as the new feathers are nearly of the same pattern and color as the old and not in contrast. . . ."

In appearance, the first winter plumage is like the previous one, "but is whiter below and richer in chestnut streakings above and below. The lateral crown stripes are distinct with black streaks, the median and superciliary stripes distinctly olive-gray. Below, white washed with pale vinaceous cinnamon on sides of head, across jugulum and on sides, and streaked, except on chin and mid-abdomen, with clove-brown bordered with chestnut, the streaks becoming confluent at sides of chin and on mid-throat forming three nearly black spots. Old and young become absolutely indistinguishable in most cases, young birds with the wing edgings perhaps a trifle duller and with a yellowish tinge." In this plumage females are "apt to be more washed with brown or to have a yellowish cast when compared with males" in the same plumage.

The first nuptial plumage is acquired, according to Dwight, by wear, which is marked; "by the end of the breeding season the birds are in tatters. The buff is lost and the streaking below comes out in strong contrast on a white ground." The adult winter plumage is "acquired by a complete postnuptial moult beginning usually about the middle of August and completed before the end of September. Old and young cannot be told apart with any certainty, adults however with wing edgings that may perhaps average darker and browner and the throat markings blacker."

The adult nuptial plumage is "acquired by wear as in the young birds with the same results."

The following description of the adult plumage is by Robert Ridgway (1901):

Adults (sexes alike):--Pileum brown (mummy brown to almost burnt umber), narrowly streaked with black and divided by a narrow median stripe of gray, more or less streaked or washed with brown; scapulars and interscapulars black medially, producing streaks of greater or less width, these margined laterally with brown (like the color of the pileum), the edges of the rectrices, more or less broadly, brownish gray; rump olive-grayish, more or less streaked with brown (sometimes with blackish also); upper tail-coverts browner than rump and more distinctly streaked; tail brown (broccoli brown to russet brown), the middle pair of rectrices with a narrow median stripe of dusky brown, the inner webs of the other rectrices darker brown than outer webs; lesser wing-coverts brown; middle coverts brown, margined terminally with pale brownish gray, and marked with a more or less distinct median streak or spot of dusky; greater coverts brown, margined terminally with paler and marked with a broad median tear-shaped (mostly concealed) space of blackish; tertials mostly blackish, but outer webs chiefly brown, passing into a paler (sometimes pale grayish or almost grayish white) hue terminally; rest of remiges dusky, edged with paler or more grayish brown; edge of wing white; a broad superciliary stripe of olive-gray, sometimes approaching grayish white on the lower portion; loral, suborbital, and auricular regions darker olive-grayish, the latter margined above and below by narrow postocular and rictal stripes of brown, these brown stripes sometimes narrowly streaked with black; a broad malar stripe of dull white or pale buffy, margined below by a conspicuous submalar stripe or triangular spot of black or mixed brown and black; under parts white, the chest marked with wedge-shaped streaks of black, more or less broadly edged with rusty brown, these streaks more or less coalesced in the lower central portion of the chest, or upper breast, forming a more or less conspicuous irregular spot; sides and flanks streaked with black and rusty brown, the ground color, especially on flanks, more or less tinged with pale olive-grayish or buffy; under tail-coverts white of pale buffy, more or less streaked with brown; maxilla dusky brown, paler on tomia; mandible horn color; iris brown; tarsi pale brown, toes darker.

Albinism occurs in song sparrows, and Root (1944) reports a banded individual that acquired a considerable degree of whiteness during a 28-day period in early autumn, presumably as the result of a molt.

Food.--S. D. Judd (1901) has described the diet of song sparrows without regard to race. For the year, animal matter constitutes 34 percent of the total food, the greatest amount being taken from May to August, when insects represent about half the bird's food. Ground-, leaf-, and clickbeetles, weevils, and other beetles rank first in number; grasshoppers, locusts, larvae such as the cutworm and army-worm, ants, wasps, ichneumon flies, bugs, leaf-hoppers, larvae and imagos of horse-flies, etc., are also taken. The remaining two-thirds of the diet is composed of seeds of crabgrass and pigeon-grass, timothy, old-witch grass, barnyard grass, panic-grasses, orchard and yard grasses; knotweeds, wild sunflower, lamb's quarters, gromwell, purslane, amaranth, dandelion, chickweed, dock, ragweed, sheep-sorrel and wood-sorrel; a little grain, largely waste; and, before the seeds have ripened, wild berries and fruits such as blackberries, strawberries, blueberries, elderberries, and raspberries, wild cherries and grapes, and woodbine berries. The species is very beneficial as a destroyer of injurious insects and weed seeds. Judd says, "Only 2 percent of the food consists of useful insects, while 18 percent is composed of injurious insects; grain, largely waste, amounts to only 4 percent, while the seeds of various species of weeds constitute 50 percent."

W. L. Dawson (1923), writing of song sparrows of no specified race, says that a bird "sometimes seizes and devours small minnows."

Behavior when foraging reflects the seasonal dietary preferences already described. Eaton (1914) states that in summer song sparrows cease to feed largely on the ground and sometimes forage for insects among foliage as high as 20 and 30 feet, although usually among bushes and grass. These birds scratch the ground by kicking simultaneously with both feet. Charles H. Blake writes of watching a song sparrow catch winged termites as they emerged from their subterranean colony in early June.

Behavior.--Song sparrows are often furtive in manner, and Knight (1908) gives a good description of the behavior of alarmed birds. He states that they prefer to "work downward into the bushes with bobbing tail, hopping along from twig to twig, or skulking through the underbrush, grass and leaves. They do not fly, save from bush to bush, unless closely pursued with evident intention to flush them or do them harm." Witmer Stone (1937) speaks of "how well adapted [song sparrows] are for a terrestrial life and how rapidly they can run, mouse-like, through the grass."

Despite their sometimes secretive behavior, birds dwelling near humans often develop considerable tameness. Forbush (1929) states that they may be conditioned by feeding to come when called and tells of one song sparrow that learned to associate the sound of a bell with the fact that food was to be scattered and of another that learned to peck at a window to be fed.

The behavior of females on the nest is often cryptic, according to Knight (1908). Sometimes the bird sits until almost stepped on; at other times, when the male gives the alarm, she slips off, sneaks a few feet away, and then begins to call. Johnston (1957) in long experience with the shrub-nesting race, samuelis, saw only one case of rodent-like distraction display. Both adults protest intrusions into the vicinity of the nest, and Forbush (1929) describes the posture of nest defense as involving "outspread wings and depressed tails." This threat, if unsuccessful, may be replaced by attack. Birds as large as the catbird and hairy woodpecker happening to approach the nest are attacked, and Forbush mentions a successful attempt by a song sparrow to drive five house sparrows from a feeding station.

Bathing, states Forbush, occurs "during the day whenever opportunity offers" and, if there is water, "every night after sunset." Puddles, including salt water along the shore, are used; and the song sparrow is one of many species that bathe in drops of water on grass and leaves by striking the foliage with the wings and body and thus throwing water on the plumage. Scratching of the head by song sparrows is "indirect," with the foot brought over the wing to reach the head, and Hailman (1959) has observed song sparrows and other emberizines scratching the head against perches.

Anting has often been noted in song sparrows; Whitaker (1957) has summarized the details.

"Helping," i.e., the feeding of young both of other song sparrows and of other species, has been noted several times and summarized by Brackbill (1952) and Skutch (1961). Perhaps the most surprising instance, reported by Brackbill, involved the cooperative building and joint use of a nest by a pair of cardinals and a pair of song sparrows. Both females incubated, the cardinal sometimes sitting on the sparrow. The cardinal eggs succeeded, and all four adults fed the nestlings. Forbush (1929) describes an instance in which two females laid a total of eight eggs in one nest; one of these birds had its own nest 30 feet away but did not use it. The females took turns incubating, and all eggs were said to have produced fledglings.

The flight speed of song sparrows has been measured by O. P. Pearson (1961) as 15.9 miles per hour and possibly as much as 21 miles per hour.

Manwell and Herman (1935) found that individual song sparrows displaced and released in spring as much as 1 1/2 miles returned quickly to the point of capture; at an intermediate trapping station on the presumed line of flight none were caught.

Body temperatures of 64 individuals of the race euphonia averaged about 109.6^o F., varying 10^o F. within the sample (Becker and Stack, 1944). There is considerable literature on song sparrow weights; Mrs. K. B. Wetherbee (1934) gives many data. Mrs. Nice (1935) lists an average weight of 21.3 grams for 267 adults. LeRoy C. Stegeman (1955) reports weight fluctuation amounts at times to 20 percent in 24 hours, with peak weights in the spring recorded in the late morning and late afternoon.

Voice.--Male song sparrows sing their variable repertoire not only during the breeding season but at other times. Songs may be heard in much of the breeding range during any month of the year (Saunders, 1947), and dawn singing on clear, cold mornings in January and February is especially noticeable. Regular singing in spring begins in Connecticut in late February or in March (Saunders, 1947) and generally closes, for a time, in the third week of August (Saunders, 1948a). There follows a revival of song, beginning in Connecticut on the average date of September 30 and continuing until November 21 (average). Aretas A. Saunders, who is responsible for these dates, writes (1948b): "This species is the most regular and dependable fall singer of all our birds."

Forbush (1929) states that the birds sing "no matter how very stormy the weather" and sometimes even "in the darkness of night." Six to eight songs per minute is the frequency during the dawn hours of the breeding period (Forbush), with singing continuing, but less frequently, all day. Knight (1908) reports that in summer in Maine most singing occurs during the dawn and evening hours. Forbush says that occasionally molting birds sing a whisper song.

When singing, the male mounts to a position typically between 7 and 15 feet from the ground (Eaton, 1914); trees, shrubs, fences and boulders are among the song posts used. Singing has often been observed in birds on the wing, and Forbush's description, quoted under Territory, indicates that this form of behavior may have become an element in some displays.

Female song sparrows have been known to sing, and Mrs. K. B.Wetherbee (1935) has written of a female in Massachusetts that sang "a clear series of whistled notes" from April to mid-June.

Not only the vociferousness of the species but also the unusual variability of the repertoires of the individual males are responsible for an abundant descriptive literature. Further, the use of electronic recording and analysis of songs has thrown light on the extent to which song sparrow vocalizations are modifiable as opposed to innate. For the present account, two authorities on bird song are quoted. Aretas A. Saunders wrote Mr. Bent: "I have 885 records of the song, no two of them alike. If we count trills as single notes, the number of notes per song varies from 4 to 20, averaging about 11. The length of songs varies from 1.8 to 5.2 seconds, the average being 2.7. The pitch varies from D'' to F''''. The pitch interval varies from 1 to 7 1/2 tones, the average about 3 1/2 tones. Each individual song sparrow sings a number of different songs. It commonly sings the same song over a half a dozen times or so, and then takes up a different song. The number of songs per individual varies from 6 to 24, the latter being an unusual bird."

The same author writing elsewhere (1951b) goes into additional detail and reports the following: Pitch varies from 1150 to 5450 vibrations, in notes audible to man, and pitch intervals are similar to those in human music. There is little variation in intensity. "Quality is usually sweet and musical. . . . Consonant sounds are not very noticeable. . . . The song has three parts: strongly rhythmic introductory notes, a central trill, and a final series of rather irregular and indefinite notes. . . . Songs are of five types. . .[differing] primarily in the position and relative pitches of the introductory notes and the trill." Donald J. Borror (1961), who analyzed 889 tape recordings of songs from 113 different individual song sparrows of the races melodia and euphonia, writes:

The songs of this sparrow consist of a series of different phrases (mostly 1- to 4-noted), and usually a trill; many of the notes are buzzy. . . . A given bird has a vocabulary of a large number of notes and phrases, and these are variously combined to produce up to a dozen or more different song patterns; the different patterns of a given bird are often quite different. The songs of a given pattern may vary. . . .

Song Sparrow songs are of two general types, those beginning with two to four (rarely one or five) similar and equally spaced phrases, and those beginning with four to twenty similar phrases that increase in tempo. . . .Songs of the first type were much more common, making up 83.8 percent of the Ohio [euphonia] patterns and 86.7 percent of the Maine [melodia] patterns. . . .

A Song Sparrow apparently has an inborn tendency to sing songs of two general types, but it learns its phrases by listening to other, nearby Song Sparrows. As a result, the songs of different birds in a local population contain similar notes and phrases (but usually arranged differently), while the songs of birds in separated populations contain different phrases. The farther away two populations are, the less likely they are to use similar phrases in their songs.

In a later, very detailed analysis of variation in the songs of Maine song sparrows, Borror (1965) found 544 song patterns represented in 7,212 tape-recorded songs of 120 birds.

A description of autumnal song by Forbush (1929) probably is applicable not to adults but only to birds of the year: "Most of the singing [in fall] is quite different, ranging from a low connected warble to a song resembling that of the Purple Finch, and (rarely) one like that of the Vesper Sparrow. There is a particularly low, sweet melancholy warble uttered just before the bird departs for the south." Formless, continuous warbling of the kind described is commonly a stage in the ontogeny of song in passerines (Lanyon, 1960).

Call notes are described as "tchenk," "tchip," "tchunk," "chip," "tcheek," "chuck." Forbush also mentions a note "sst"; a similar note given by California races is regarded by Dawson (1923) as functioning as a flocking or recognition call.

Field marks.--A medium-sized sparrow, the song sparrow is best recognized by the heavily streaked breast, on which the streaks are "confluent into a large central spot" (Peterson, 1947). In flight which is usually for short distances between perches or into cover, the bird is distinguished by its manner of pumping its rather long, rounded tail up and down. Witmer Stone (1937) describes this flight graphically as " 'brokenbacked'. . .as if the tail were hinged at the base."

Banding.--The longevity record for banded song sparrows appears to be about 8 or 9 years. Mrs. Nice reports a male that was at least 7 1/2 and possibly 9 1/2 years old at death. A song sparrow Mrs. K. C. Harding (1943) banded April 27, 1936, at Cohasset, Mass., and recaptured there on April 5, 1943, was in at least its 8th year.

Recaptures and recoveries at points other than the original banding station have been reported with some frequency in the journal Bird-Banding. Some of the most interesting of these follow: The number of such "recoveries" and "foreign retraps" from a total of 3,614 song sparrows banded in Montgomery County, Pennsylvania, was 12 (Middleton, 1956); from 6,109 banded in Groton, Mass., 7 (Wharton, 1953); from over 1,200 banded on Cape Cod, Massachusetts, 1 (Broun, 1933). Some of the Pennsylvania birds were caught in Georgia, North Carolina, Maryland, and New Jersey. The Groton, Mass., birds were caught in Arkansas, South Carolina, North Carolina, Nova Scotia, and New Brunswick. The Cape Cod bird was caught in South Carolina. May Thacher Cooke (1943) reports a bird banded in Massachusetts and captured in Newfoundland. Wendell P. Smith (1942) caught a bird in Vermont in April, and it was recaptured 90 miles southward in New Hampshire in June of the same year, an interesting case of reversal in the direction of migratory movement.

Enemies.--Perhaps the most interesting and surely the most well documented hazard in the song sparrow's environment is exposure to the parasitic brown-headed cowbird and bronze cowbird. Herbert Friedmann's latest report (1963) on the cowbirds states that the song sparrow (all races) shares with the yellow warbler the claim to being the most frequently reported host of M. ater. Friedmann's summary of the relations between all races of the song sparrow and the brown-headed cowbird is quoted substantially in full:

The song sparrow is one of the most frequent, if not the most frequent, victims of the brown-headed cowbird. Since the former is sympatric with the latter throughout the entire breeding range of the parasite, it is parasitized probably more often and over a greater area than any other bird. The total number of records is very great. After accumulating over 900, I stopped noting them except for records of special interest. The data came from every province of Canada and every state of the United States included in the breeding ranges of both birds. All three races of the parasite are involved, and no less than 17 races of the song sparrow: melodia, atlantica, euphonia, juddi, montana, inexpectata, merrilli, fisherella, morphna, cleonensis, gouldii, samuelis, pusillula, heermanni, cooperi, fallax, and saltonis. So far, none of the purely Mexican races have been reported as fosterers of the cowbird, but this fact is probably due more to a lack of human observation than to any actual immunity of the bird to cowbird parasitism.

There is no need to detail actual instances for the various races of the song sparrow since such cases have already been given in my earlier summaries [Friedmann, 1929, 1934, 1938, 1943, 1949]. However a few additional records of infrequently reported races of the host species should be mentioned. . .[viz. cleonensis, fallax, morphna, saltonis, and inexpectata].

In recent years, not only many hundreds of additional cases, but also much more quantitative data on the host-parasite relations have become available. Hicks (1934) found that 135 out of 398 nests (34 percent) of this sparrow were parasitized in Ohio. Nice (1937a, 1937b), also in Ohio, reported that 98 out of 223 nests (43.9 percent) contained eggs or young of the cowbird (the annual percentage varied from 24.6 to 77.7 percent). Sixty-six unparasitized nests raised an average of 3.4 song sparrows, whereas 28 successful but parasitized broods averaged only 2.4 song sparrows, indicating that each cowbird was reared at the expense of one song sparrow. In one instance Nice found that a pair of song sparrows raised a young cowbird together with five of their own young. Apparently here no loss of sparrows was involved. In another paper, Nice (1936) noted that, in all the song sparrow nests which she had watched during a period of five years, adult cowbirds removed 5.7 percent of the song sparrow eggs and nestling cowbirds crushed or starved 3.5 percent of the young sparrows. The cowbird eggs did not succeed as well as those of the host; only 30.7 percent of the former, but 35.8 percent of the latter, reached fledgling stage. In 1930 - 31, there was one female cowbird to about 11.5 pairs of suitable hosts, but in 1934 - 35 there was one to 8.6 pairs of suitable victims.

Of all song sparrow nests parasitized, Nice reported that 70 percent held a single cowbird egg each, 27 percent held 2 each, and 3 percent held 3 each. In the area of the study--near Columbus, Ohio--the song sparrow was the most important host of the paeasite [sic.]. Norris (1947) noted that 11 out of 27 nests (40.7 percent) in Pennsylvania were parasitized, and Berger (1951) recorded 37 out of 59 nests found in Michigan (62.7 percent). In the Detroit area, as reported by the Detroit Audubon Society (1956), the average frequency of parasitism of the song sparrow was 40.1 percent of all the nests found. . . .

One is drawn toward attempting an over-all estimate of the frequency with which the song sparrow is victimized, but to do so with any feeling of accuracy is difficult because the incidence of parasitism appears to vary geographically (or, at least, the frequency with which it is reported varies). From this it follows that the over-all percentage depends on how many geographically different areal data are used in the estimation. [With one group of studies of the eastern United States], we come up with a total of 323 parasitized nests out of 804 nests observed, or a little over 40 percent. On the other hand, in southern Quebec (Terrill, 1961, p. 11), out of 486 nests observed, only 62, or 12.7 percent, were parasitized. If we put all these studies together, we get a total of 382 out of 1,285 nests victimized, or 29 percent. This figure becomes yet smaller when we attempt to include data from other parts of the continent.

[One color-banded song sparrow in a single summer] had no fewer than five consecutive nests. . . . It would seem that, if none of these nests had been interfered with, there would not have been sufficient time for four or five in one season. . . . It appears that one of the effects of parasitism may be to increase the "nesting potential" of the host. . . .

As many as 7 cowbird eggs have been found in a single nest of this sparrow; there are numerous records of 3, 4, and 5 parasitic eggs to a nest. Occasionally, but not often, song sparrows may partly bury cowbird eggs by building a new nest lining over them--if the alien egg is laid before any eggs of the host.

Salmon (1933, p.100) has reported seeing a song sparrow feeding three fledgling cowbirds; no young sparrows were mentioned. Lees (1939, p. 121) recorded that near Wetaskiwin, Alberta, he watched a song sparrow feeding no less than five young cowbirds. This must be a record of fledgling success for any host species.

Friedmann (1963) lists no instances of parasitization by the bronze cowbird of subspecies of the song sparrow covered by these life histories. Only the Mexican race mexicana has been involved.

Other enemies of the song sparrow are at least four species of hawks (Munro, 1940; Randall, 1940; Hamerstrom and Hamerstrom, 1951; Heintzelman, 1964) and at least five species of owls (Allen, 1924; Hawbecker, 1945; Johnston, 1956; Fisler, 1960; Graber, 1962). S. A. Altmann (1956) has reported mobbing of mounted specimens of both screech and great horned owls, and F. Hamerstrom (1957) witnessed mobbing of a tame red-tailed hawk. Forbush (1929) mentions nest defense against snakes and turtles; he does not indicate what turtles may be involved, but box turtles of the middle west (Terrapene o. ornata), at least, have been known to eat birds and their eggs (Legler, 1960).

A curious case in which a garter snake (Thamnophis s. sirtalis) disgorged an adult song sparrow is reported by Carpenter (1951), who suggests that the bird must have been found dead and then eaten. Mahan (1956) discovered a milk snake (Lampropeltis triangulum) eating song sparrow eggs in a nest 15 inches above the ground.

A number of external parasites taken from song sparrows east of the Mississippi River, most of them within the range of melodia, have been reported by Harold S. Peters (1936). These include the Mallophaga Degeeriella vulgata (Kell.), Machaerilaemus maestum (Kell. and Chap.), Menacanthus incerta (Kell.), Philopterus subflavescens (Geof.), Ricinus melospizae (McGregor); the bloodsucking hippoboscid flies Ornithoica confluenta Say, and Ornithomyia anchineuria Speiser (syn. O. fringillina); the mites Analgopsis sp., Liponyssus sylvarium (C. and F.), Trombicula bisignata Ewing, Trombicula cavicola Ewing; and the ticks Haemaphysalis leporispalustris Packard, Ixodes brunneus Koch, and Ixodes sp. Herman (1937) gives further data on the hippoboscids parasitizing song sparrows, as does Boyd (1951); the latter would apparently refer the records of Ornithoica confluenta to Ornithoica vicina, as the parasite of song sparrows.

Nestling song sparrows are among the many species victimized by the maggots of blow flies (Calliphoridae). Johnson (1932) found larval Protocalliphora splendida (Macq.) in a nest, and George and Mitchell (1948) report Apaulina metallica (Townsend) (syn. Protocalliphora metallica).

Blood protozoa found in song sparrows (Herman, 1944) include a number of species of the genera Haemoproteus, Leucocytozoon, Plasmodium, Toxoplasma, and Trypanosoma.

Cats, other predatory mammals, and man are often responsible for song sparrow deaths; and the physical environment takes its toll in starvation (Forbush, 1929) and in the flooding of ground nests.

Song sparrows are among the birds whose feet occasionally exhibit large, rough, wart-like swellings. H. and J. R. Michener (1936) have described the appearance and effect of this disease, which they also imply may affect the wings and heads of song sparrows, and which they regard as mildly contagious and epidemic. Their observations were made in California. The disease runs its course between from 1 to 5 months. Apparently it often produces no noticeable after-effects, but it may cause the loss of the nails and sometimes the phalanges. These authors quote a pathologist's histological analysis of a foot of an unspecified species. The opinion was that the lesions were not true tumors but were the result of an unrecognized irritant or infection. Viruses are now known to produce comparable effects in some birds (Herman, 1955).

Fall.--In Massachusetts, a few song sparrows begin moving away from their breeding places in mid-July. The fact that nests have been found in New York as late as August 25 (Eaton, 1914) suggests that it may be the young of the year that move at so early a date. There is some conflict in the evidence as to when the fall migration is at its peak. The data of Steward and Robbins (1958) are perhaps the most recently reported and based upon the most voluminous and varied factual data. These authors state that in Maryland and the District of Columbia the normal period of fall migration is between Sept. 20 - 30 and Nov. 20 - 30, with its peak between October 10 and 30. More northerly latitudes would, of course, be correspondingly earlier, e.g., in Maine "throughout September and most of October" (Palmer, 1949).

Winter.--Most song sparrows that pass the winter in Massachusetts (Forbush, 1929) remain near the sea, where there are usually patches of ground clear of snow. In Ontario (Snyder, 1951) and New York (Eaton, 1914) the habitat at this season is principally marshes and swamps. In Pennsylvania, Todd (1940) says brushy thickets and fields with corn shocks are frequented by song sparrows. The birds are not especially social, but they are often seen in loose flocks of mixed composition and, particularly in severe weather, may assemble in small companies with other song sparrows.

South of the breeding range, in the deep south and southeast of the United States, melodia is found with euphonia, juddi, and, in places, atlantica. Here the birds seek the same brushy, moist, riparian and marshy situations that they prefer for breeding. Sprunt and Chamberlain (1949) describe such habitat in South Carolina and say the song sparrow "often. . .is found with swamp sparrows. . . . The thoroughly characteristic song is delivered throughout the winter except in very cold weather or on freezing days." In contradiction Arthur H. Howell (1932) in his work on Florida says, "The Song Sparrow, so well known in the North by its cheery song, is practically silent during its stay in the South, except for its metallic, characteristic tchip." Howell adds that the birds associate "in small loose companies, but not in compact flocks." George H. Lowery, Jr. (1960), of Louisiana, emphasizes the "entirely different personality" of the song sparrow that winters in the south and describes it as a shy and, it seems, silent skulker that prefers the depths of thickets or "a rank growth of broom sedge."