VoL XLII] 1925 I MILLER, The Mandarin and WoodDucks. 41 with insects,constitute the principalfare during early life. But as the grow, rats assumea more importantrole, and in or neartx;•o different nests I foundremains, picked clean, of practically full-grownAmerican Bitterns (Botauruslentiglnosus). Now the youngAmerican Bittern is no meanantagonist, and the fact that suchlarge birds are actually killed and carriedto the nestindicates the calibre of the Harrier as a hunter. I recall, once before in the later fall, flushinga Harrier from the half-consumedremains of an adult Green Heron. This had possiblyacquired a taste for Ardeidae in earlier life. Duringthe fourth week of the young Harrier's life pellets of fur and feathers,containing some bone, begin to appear about the nest. These pellets are often as large, as compact and as well formedas thoseof the Short-caredOwls, constituting an interesting similaritybetween the two . It is probablethat the failure to find pelletsabout the nestsearlier in the youngbrood's growth is due to the thoroughremoval of wasteby the adults,rather than any changein feedinghabits. 613 ClsvelandAve., Elizabeth,N.J.

THE SECONDARY REMIGES AND COVERTS IN THE MANDARIN AND WOOD .

BY W. DE W. MILLER.

T•E most striking peculiarityof the male Mandarin is ßthe remarkablesail- or fandikeexpansion of the twelfthsecondary, an ornamentunique among birds. Correlatedwith this are other equally unusualfeatures, evident only on closeinspection. The thirteenth secondaryis reducedto a mere hidden vestige much shorterthan its greatercoverts, and the fourteenth,though twice a s long, is of far lessthan normalsize. Comparisonof the wingwith that of the WoodDuck ( sponsa) provesof greatinterest. In the latter there are severalremarkable features,notably the abbreviationof the twelfth secondarywhich is only four-fifthsas long as the remex on eachside of it. The peculiaritiesin the remigesof this specieswere pointed out by Mr. Proximal Secondariesof left wing from above, the coverts removed: A, of (Aix spons•),adult male; B, of Mandarin Duck (Dendronessa galericulata),adult male. In eachcase the abbreviatedsecondary (fourth from inner edge) has been raised. Vol.1925 XLII] ] MILLER,The Mandarin and Wood Ducks. 43

'Ridgway many years ago (Water Birds of North America, II, p. 12, 1884), but I can find no mention of the reducedsecondaries in the Mandarin nor any comparisonof the remigesof the two species,and it thereforeseems worth while to figure and describe them. Furthermore,the female Wood Duck has a peeullarity of its own in the greater secondarycoverts. In order to correctly comparethe abnormalremiges of the two speciesit is essentialthat their homologlesbe detm•nined. For this purposewe may first examine the undifferentiatedwings of the females. We find in the Mandarin eleventypical secondary remigesand five inner secondariesor "tertials," whleh differ in size, shapeand colorfrom the others. In the Wood Duck these remigesnumber ten and five respectively,i.e., there is one less typlealsecondary than in the alliedspecies. The tertialsare plain bronzy-olive,the first one longerand broaderthan the adjacent secondary,the four others graduated. We must decidewhether there has been an abrupt lossor gain (accordingto the species)of one of the typical secondaries,or whether there has beena changeat the proximalend of the series and the transformationof a tertial into a typical secondaryor vice versa. The longesttertlal is so unlike the adjacent typleal secondary that I believe any modlfieationof one into the other •Sthln the limits of the genusis most unlikely. Moreover, I have examined onefemale Wood Duck in whichthere are eleventypical secondaries as in the Mandarin, and in a male of the latter there are eleven in one wing and twelve in the other! It is fairly safe to assumethat the first eleven secondariesof the femaleMandarin correspondto the first ten of the WoodDuck and that while the long bronzy-oliveouter tertial is the twelfth quill in the former and the eleventh in the latter, yet the two are strletly homologous. The number of secondaries in the males is the same as in the femalesand on this assumptionit is evidentthat the remarkable sail-liketwelfth remexof the Mandarin correspondsto the eleventh quill of the Wood Duck and in eachease it is the first tertial. While these feathers are merely modifiedinner secondaries,it will be more convenientin the presentpaper to refer to them as 44 MILLr,R• The Mandarin and Wood Ducks. [J•n.[Auk tertials. To l•lr. JamesP. ChapinI am greatlyindebted for the two accompanyingfigures.

COMPARISON OF THE SECONDARIESIN THE MALES. The first (distal) nine secondariesare normal in size and shape and virtually alike in the two speciesthough broader in the Wood Duck. In the Mandarin they are bronzy-olive,the last oneor two being steel-blueon the outer web, while in the Wood Duck all are steel-blue. In both specieseach feather is narrowly tipped with white. The next two feathers in the Mandarin are blue like the ninth but lack the white tip, or the tenth may have a very narrowtip. In the WoodDuck the tenth feather,which probably correspondsto the eleventhof the Mandarin and like it always lacksthe whitetip, differsin havingthe outerweb, which is largely bright copperyred, widening terminally, and the endof the feather angular. The twelfthquill or sail-featherof the Mandarinis, as already stated,the first tertial, and homologouswith the eleventhof the Wood Duck. In both birds,but particularlyin the Mandarin, this remexis muchlarger than the precedingfeather and widenstowards the end, but while in the Mandarin only the inner web is expanded, the outer beingeven narrower than usual,in the Wood Duck the outer web is considerablywider than the inner. In the Oriental bird the outer web of this sail-likefeather is violet-bluenearly to the tip, the rest of the feathertawny-rufous with blackand white terminal edges. In the Americanbird the inner web is fuscous, the outer velvety black, showingrich bluish greenreflections in certain lights, the tip of the featherpure white. The second tertial of the Mandarin is a miniature of the first, but the shaft is straight and the tawny inner web degeneratein character. It is a mere vestige ranging from fourteen to thirty millimetersin length(usually about 17) in the few birdsexamined. The outer web is violet or this colormay be moreor lessreplaced by white or fuscous. It is crowdedin betweenthe neighboring quillsand hidden by theseand by its uppercovert which is two or three times as long. Except that the correspondingfeather of the Wood Duck is Vol. XLII] 1925 I MILLER, The Mandarin and WoodDucks. 45 considerablyshorter than the immediatelyadjacent tertials there is no resemblancebetween it and the Mandarin's vestigialremex. It resemblesits precedingtertial in colorand form but lacksthe white tip. Thethird tertial Of the Mandarin, while strikingly abbreviated, is about twice as long as the second. It is obliquelycut off on the inner web, the outer web tapering to an obtusepoint; light violet blue, the inner web sometimeswith a narrow white terminal edge. In the Americanspecies this feather is whollydifferent and dosely resemblesthe first and secondtertials, being of the same length as the first but, like the second,with no white tip. The last two tertials of the Mandarin are plain bronzy-olive feathers,the fourth a little shorterthan the third, the fifth only half as long as the fourth. In theWood Duck the fourth tertial is about three-fourthsto four-fifths as long as the large third one. It resemblesthe latter in colorbut is usually duller and is much narrower and does not widen terminally. The last tertial is a small plain feather, sometimesdegenerate and shorter than its covert, or twice as long, (considerablyexceeding its covert) and firm webbed. The anomalous relations of the second and third tertials of the male Wood Duck is reflected in the female in which the third feather is often scarcelyshorter than the second. In the female Mandarin there is no suchapproach to the charactersof the male but the first tertial is much longerand broader than the others. To sum up, the first three tertials are conspicuouslymodified in the malesof both species,the first two at least are broad and truncate, and the secondone shorter than the first and third. All three of these feathers, however, are much more modified and peculiarin the Mandarin than in theBrood Duck and the difference in size between them is far more marked. Thus while the first tertial of the former is more expandedthan in any other Duck, the next one is unique in its vestigial condition. Furthermore, in the Mandarin the inner web of each feather is far wider than the outer, whereas in the Wood Duck the outer is the wider. The remigesimmediately adjoining these tertials also differ con- spicuouslyin the two species. In the male Wood Duck the first and third tertials are equal 46 MILLER,The Mandarin and Wood Ducks. [Jan.[Auk and longest,only the last t•vo being graduated, while in all other membersof the family at least three are shortened. In the •VIan- darin,on the otherhand, the third is disproportionatelyabbreviated and actually much shorter than in any other Anserincbird. As already stated, the Wood Duck has but fifteen secondaries, one fewer than the Mandarin. This is the minimum number in the , for of the thirty-six genera examined by me none has fewer than sixteen secondaries,and all but six have more than sixteen. In certaingroups of birds,as the Tyrannidaeand Cotingldae,the primaries are not infrequently irregular in length, and this is carried to an extremein Tityra in which the ninth primary is reducedto a vestigebetween two quills of normal length. So far as I am aware no suchreduction in the length of a sec- ondaryhas heretofore been recorded in any bird, for whichreason the state of thesequills in the Mandarin and Wood Ducksis especi- ally notable. In •VIr. Ridgway's note on the male Wood Duck, above cited, he describesthe broader, black scapularsas alternating with narrower,more poin•d, bronzy ones,which are mostly concealed betweenthe broaderfeathers. In the freshspecimen now before me I find that the black feathers are the outer ones, the bronzy feathersthe inner ones. There is no alternation as regardstheir insertion in the skin but in the normal contracted condition of the humeral tract the feathers naturally overlap in this alternate fashion. A differencein the greaterupper coverts may alsobe mentioned. In the •VIandarin these coverts are of the same bronzy-ollve as the other upper covertsand the innermosttwo are graduated. In theWoodDuck the outer covertsare olivebut as they approach the elbow the coverts gradually become green and then violet, with a natrow black tip, and at the sametime becomeactually shorter and broader. The penultimate one, however, is green with no black tip and is slightly longerand narrowerthan the one distad of it. Thus only the final covert is shortened. In the femaleWood Duck there is a very interesting peculiarity in the uppergreater secondary coverts. The four of thesecoverts belongingto the tenth to thirteenth secondariesare largely bright Vol.1925 XLII] I M•LLEn, The Mandarin and WoodDucks. 47 red-purplecontrasting conspicuously with the adjacentsteel-blue and olive-brown. Their chief distinction, however, is that they are much larger than the precedingcoverts, particularly that of the eleventhquill, which is considerablylonger as well as broader than the more distal coverts. The two covertsproximad of this one are graduated and the one distad of it is also shortened. There is no such modification either in size or color in the male Wood Duck nor in either sex of the Mandarin. In these the covert of the eleventh remex is smaller than the distal coverts instead of larger.

OTHER DIFFERENCES RETWEEN TItE SPECIES.

The Mandarin is a slightlysmaller bird with a decidedlysmaller bill, but the tarsi are actually longer than in the Wood Duck and the wings are fully as long. There is no backward extensionof the bill in a point on each side of the foreheadas is so conspicuouslythe easein the Wood Duck, and the teeth of the maxillarytomia are muchless developed and usually invisiblefrom the side. The very long, narrowly linear~lanceolate"hackles" which springabundantly from the posteriorpart of the headand anterior part of the neck in the Mandarin (shorterones even crossingthe fore-neck)are entirely different from the short velvety of the Wood Duck. In the Mandarin there is a graduated outer seriesof falcate, pointed scapularfeathers, pure white with a broad deep black outeredge (much like thoseof Nettioncrecca). In the alliedspecies thesescapulars are very broad,rounded terminally, somewhat arched or convex and velvety black with violet reflections. The central one or two pairs of rectriees in the Wood Duck are broader, rounder at the end, and laterally arched or vaulted instead of flat as in the Mandarin. The upper tail-coverts too, in the latter, are flatter and narrower and there is no special developmentof the outer ones. In the Wood Duck the several outer covertsare archeddownward laterally over the baseof the tail, the very broad,loose webs with a tawny shaftstripe. In the Mandarinthe uppertail-coverts are usually slightly longer than the lower, instead of the reverseas in the other species. 48 MILLER,The Mandarin and Wood Ducks. [Jan.[Auk

I cannot confirm the marked differencesin the length and graduation of the tail and in the length of the lower coverts describedby Ridgway (1. c.) and suspectthat an imperfector molting specimen.of the Mandarin was examined. In the WoodDuck the plumageof the chestis harshto the touch and under a magnifyingglass the barbsare seento be somewhat thickenedsubterminally and devoid of barbules. In the allied speciesthe plumageis soft and normal. The plumageof the upperparts in generalis richer,softer and more lustrousin the Wood Duck, violet, red-purpleand velvety blackpredominating, and the bronzy-oliveis richerthan in the Man- darin. The feathersare larger,broader, and moreobtuse, and tend to be convex or arched rather than flat. The colorand pattern of the head(except the crown)and of the bill, the scapularsand the underwing- and tail-coverts, are all very different in the two birds. In their colorationthe femalesof the two speciesdiffer chiefly as follows. In the Mandarin the upperpart of the head is grayer, the grayish olive back and scapularshave no red-purpleshade, the upper wing-covertsare entirely bronzy-olivewith noneof the steel-blueor copper-redof A. sponsa. On the outer web of the ninth secondary,and often of the eighth also, is an obliquewhite bar (lackingin oneof eightspecimens); the outerweb of the tenth quill is edgedwith white (in all but one of the eight birds). The under-wingcoverts differ as they do in the malesand the crissum also to a certain degree. The structural differencesbetween the two species,then, lie in the form of the bill and in the very different developmentor shape of the feathers of various parts, as the neck, scapulars, secondaries,upper tail-covertsand centralrectrices. The females showthese differences (except in the scapulars)to a certaindegree, and, furthermore,in this sexof the Wood Duck there is an unusual developmentof someof the secondarycoverts which are normal in the other species. The Mandarin is not only unique in the neck ruff or cape and in the form of the tertials, but the wingsare of maximumrelative lengthwhile the bill is probablysmaller than in any other fresh- water Duck. Vol. XLII] 1925 J MILLER, The Mandarin and WoodDucks. 49

The numberof rectricesin this group is sometimesgiven as sixteen. Of the thirteen Wood Ducks examined, eleven have fourteen rectrices,one fifteen and one sixteen. Of the ten Man- darins, eight have fourteen,and two have fifteen The variation is irrespectiveof sex. There has long been a differenceof opinion as to whether the WoodDuck and the Mandarin Duck are congenericor not. They are obviouslymuch more like each other than either is to any other member of the family. Indeed, so distinct are they from any other groupthat their exactaffinities are somewhatuncertain. The natural and usual coursetherefore has been to keep them togetherin the genusAix. In this comprehensivesense Aix differsfrom all other Ducks in the silvery white edgingof the primaries,the elegantblack and white terminal banding of the vermiculated flank feathers and the very broad, truncate tertials, the secondof which is smaller than the third. The lustrousgreen-and-white crest, the long wingsand the relativelylong, broad-leathered tail, with longcoverts, are alsocharacteristic. Someof thesefeatures are presentin the male only, others in both sexes. Ridgway (op. clt. p. 9) has remarkedthat the Mandarin differs from the Wood Duck "in so many pointsof externalanatomy as to render it extremely doubtful whether the two speciesshould be kept togetherin the samegenus." They are perhaps as distinct as are the monotypicgenera Charitonetta,x Arctonetta, Nomonyx and Lophodyteswhich have long been recognizedby Americanornithologists. Harterr con- sidersthe two speciesgenerically separable, although there are a number of current generaof Ducks and Geese,including Chari- tonettaand LopI•odytesthat he refusesto recognize. The Wood Duck, •daasspoasa Linnaeus, is the type of the genus Aix Bole (cf. Allen, Bull. Amer.Mus. Nat. Hist., XXIII, p. 305, and A. O. U. Check List of N. Am. Bds., 3d Ed., 1910, p. 73).

• In this connection it may be well to call a•tention to a character of Chart- tonetta that is often overIooked. According to MacGillivray (and I can con- firm his statement) the trachea has "scarcely any appearance of dilatation at the pa.rt which is so excessively enlarged in the Golden-eyed Duck." This refers to the bony labyrinth at the lower end of the trachea. 50 LINCO•.N,TheBird Life of North Dakota. [Zan.Auk

Swainson's name Dendronessa is available for the Mandarin Duck which, if consideredgenerically distinct, will be known as Den- dronessagalericulata (Linnaeus). 1 AmericanMuseum of Natural History, .

NOTES ON THE BIRD LIFE OP NORTH DAKOTA WITH PARTICULAR REFERENCE TO THE SUMMER WATERFOWL.

BY FREDERICK C. LINCOLN.

ITINERARY.

T•r fo]lowlngnotes were taken by the writer while engagedin officialwork for the BiologicalSurvey, United StatesDepartment of Agriculture,during the periodJuly 2 to August 1, 1921. The trip was made possiblethrough the cooperationof Dr. John C. Phillips,and had as its objectiveexperimental work to determine satisfactorymethods for trappingwaterfowl for bandingpurposes. I arrived at Napoleon,North Dakota, on July 5, and from that date until July 12 made observationsand conductedtrapping experimentsat lakes in the vicinity. On July 12, I went to the town of Turtle Lake, via Bismarck and Underwood, remaining there two days. Leaving that point, my next field of operations was Devils Lake, reachedby way of Minor. After examining someof the smallerlakes in the neighborhood,I selectedfor further operationswhat is known as Mission Bay, on the south side of Devils Lake, whereI campedfrom July 20 to 27, when I returned to the town of Devils Lake and concludedmy activitiesat Hank- inson on the 30th. The followingllst of lakes of the Napoleon,Hankinson, Turtle Lake, and Devils Lake regionswill simplify future referenceto the lakes and sloughsstudied:

• Most of the older authorities and even ttartert as late as 1915 give qalericu- lata as the type of Atx. Dr. Richmond informs me that these writers probably follow Gray who, in 1841, gave that speciesas the type, but that Eyton in 1838 had already designated sponsa as the type of Atx.