the Least Flycatcher and one of the Western Wood Pewee. The Goldfinches evidently did not begin nesting until late in July, just before I left for the mountains and I found no nests. During June there were often more than three pairs on the area, there being a small band of birds roving about the vicinity that were evidently not mated. By July, however, the birds had evidently chosen mates and I was able to find the three pairs regularly.

Concerning the Cowbirds, the same remarks are true here as in the cinquefoil brush association. I found more eggs of this species in the cottonwoods, but did not observe the birds themselves there so frequently. I found nine eggs of this species in nests of this association; four in nests of the Yellow Warbler, two in a nest of the Chipping Sparrow, and one each in nests of the Least Flycatcher, Cedar Waxwing and Black-headed Grosbeak.

Of the remaining six species, I found the nests of all except the Nighthawk, so have undoubtedly a correct count of them. The Chipping Sparrow and Black-headed Grosbeak are usually rare in this association, the former being common only in the evergreen forests of the mountains in this region. The latter belongs more properly to the willow brush association, where it is much commoner than in cottonwood groves. The Cedar Waxwing often occurs in greater numbers in this association than my count of one pair would indicate, but since this species varies in abundance from year to year, and is very irregular in occurrence and nesting, my count per acre would no doubt be accurate if applied to a larger area.

I am not quite sure that the Nighthawks nested in this association. I found them there all through the latter part of June and all of July, and observed both birds, resting and sleeping on the horizontal limbs of the cottonwoods. I have never found a nest of this species in this association, however, but have always found them on the prairie benches, or in the mountains along barren ridgetops. Why they frequented the cottonwoods so much unless they nested there, I cannot say, for there were no prairie benches within half a mile. These birds had certain favorite spots on the cottonwood limbs on which to rest, and after I had learned where these were I could locate both birds any day that I wished to. They were certainly mated, for their plumage showed that they were male and female, and I had observed courtship early in the season,

when another rival male was in the vicinity for a few days. I once observed a similar case of Nighthawks staying in a cottonwood grove through the breeding season, at Bozeman, Mont., and in that case saw the two young with the parents in August. I was then informed by a person whose information I had usually found to be unreliable, that Nighthawks often laid their eggs in a hollow spot on the upper surface of a horizontal cottonwood limb. Thinking this might possibly be the case with my Choteau birds, I climbed the trees, and examined the favorite perching places of these birds, but found nothing. There was a curious warfare between these Nighthawks and the Western Wood Pewees inhabiting this grove. The Nighthawks were fond of the same sort of horizontal limbs for perches that the Pewees chose for nesting sites, and whenever a Nighthawk attempted to alight on a limb near one of the Pewee's nests, the male Pewee drove it off with a ferocity that showed his relationship to the Kingbird.

Because I found certain birds nesting on the area, whose presence was due to the stream which crossed it, I found it best to add another association, the stream border. Perhaps this association could have been more easily figured in terms of length of the stream, but for the sake of uniformity in comparison with the other associations, I figured it as closely as I could in acres. This association consists mainly of the banks of the stream, a typical prairie stream, with deep pools and shallow rapids but no very swift places. It was bordered by occasional stony bars, and by steep clay banks three or four feet high. The banks were mostly grass grown, but there were a few clumps of young cottonwoods, buffalo berry bushes, and birches (Betula fontinalis). The birds found nesting along the stream border were as follows.

The Spotted Sandpipers nested on the stony bars along the edge of the stream, and I located three of the four nests. The Kill

deers evidently nested in similar places, but I did not succeed in finding the nests. There was no doubt but that the nests were there, for each pair of birds was to be found regularly in a certain spot, protesting at my presence. The Kingfisher's nest was plainly visible in the steep clay bank of the creek. The Kingfishers themselves, evidently did not live entirely within the area in question, but fished the stream up and down for a considerable distance both above and below the area.

In other parts of Montana I had always classed the Kingbird as a member of the cottonwood grove association. Here however, these birds depart from their ordinary habit, and build their nests in bushes bordering the streams. I found all four nests, and not one was at an elevation of more than four feet. All were on the bank of the stream, and two were built so that they overhung the water, one of them only about three feet above the surface, and almost on a level with the bank. This nest was beside a pool in which the boys of Choteau go swimming. Several of the boys informed me that a nest had been in that particular spot for several years. This information was supplied at different times by different individuals, so I have no doubt that it is correct.

The cattail slough association consisted of two small sloughs, with water in them about eighteen inches deep, and grown thickly with cattails, I found but two species in this association. They are as follows.

I found all five nests of the Red-wings, one in one slough, and four in the other. The actual number of adult birds was less than ten, for the Red-wings showed evident signs of polygamy. The single nest in the first slough was guarded by a pair of birds, but the other four nests, though I saw all four females, evidently had but two males.

Whether the Bittern nested or not I could not be quite certain. I saw but one bird, which pumped every evening from one or the

other of the two sloughs, or occasionally from a third slough outside the area. I rather suspected that there was a Bittern's nest in the larger slough, with the four Red-wings, but I failed to find it.

Summing up the five different associations, we have 112 pairs of birds nesting on the area of 265 acres, making an average of 2.36 acres per pair. This however does not cover all of the birds found on the area during that time. Others went there occasionally and often stayed for several days. There were of course, many migrants in spring, and some of these remained during the first half of June. Other birds, evidently not migrants, that occurred more or less frequently, are as follows. A male Red-eyed Vireo stayed in the cottonwoods and sang for several days, late in June. He was not there long enough to have had a mate and nested. A Warbling Vireo that had a nest somewhere in the shade trees in Choteau, occasionally visited the cottonwoods. A Cooper's Hawk appeared one evening in July. I found it devouring the remains of a young Brewer's Blackbird. On May 25 a pair of Bronzed Grackles appeared. They were evidently mated and might have nested on the area, but I had not at that time determined to make an ecological study, and believing that they would jeopardize the safety of other nesting birds on the area, I collected them both. This is the only pair of these birds I have ever seen in this region, which must be at about the western limit of their range.

There may be some question concerning the value of such a study as I have presented. The figures given represent only the conditions on a small area, and more extensive studies will be necessary to prove whether or not they will hold true if applied to a larger tract. I hope to make further studies along this line, provided the opportunity is again presented. A census of breeding birds could probably be most accurately obtained by studies of associations on an acreage basis, supplemented by studies of a whole region, which would give the proportion of the area of the region covered by each association. The methods of work which I have used will probably be of some value to others who wish to make similar studies in the future, but they can undoubtedly be improved upon.

DOES A GREBE SPREAD ITS WINGS JUST BEFORE

DIVING?

BY VERDI BURTCH.

Plates XXI-XXII.

ON March 10, 1912, I got a remarkable photograph of a Holboll's Grebe (Colymbus holballi) in the act of diving. It had been a very cold winter and Keuka Lake was frozen over early in January with the exception of the channel through the bar at Branchport, N. Y. Here there is a continual current flowing into and out of the harbor and this keeps the channel open even in the most severe weather. This open area is usually about ten to fifteen rods in diameter and here about six hundred wild ducks, Canvas-backs, American Scaups, American Golden-eyes and a single Buffle-head passed through the same ordeal as described by Alvin R. Cann in 'The Auk' for 1912, (Vol. XXIX, p. 437), while the conditions were the same as prevailed on Cayuga Lake where his observations were made. At the time of which I write the sole survivor of the entire flock of ducks was the little Buffle-head. We had just had several days of very severe cold and the open area was much restricted. This morning I found a Holboll's Grebe with the Buffle-head. It dived as soon as I went near and when it came up uttered a "pup-pupput" and dived again. It would stay up but a moment between the dives and sometimes stayed under water so long that I began to fear that it had come up under the ice but it eventually came up in the open area each time. As it was above water only a moment each time it was rather difficult to focus the camera and make an exposure before it dived again. Once when it came up quite close to me, I focused and released the shutter at the very instant it started to dive. The resulting picture shows a dark area representing the depression in the water where the Grebe sat at the beginning of the exposure, the 'ghost' of its wings fully spread out behind as it started to lunge forward and the Grebe itself nearly submerged a full length ahead. It would seem to me that the wings were instantly spread and closed again as the Grebe lunged forward, the