Bird Migration in the Mackenzie Valley

442 COOKE,Bird Migration in theMackenzie Valley. [Auk[oc;.

BIRD MIGRATION IN THE MACKENZIE VALLEY.

BY WELLS W. COOKE.

ThE MackenzieValley in northernCanada presentsa broad sur- face with a gentle sloperising only eight hundredfeet in the fifteen hundred miles from the mouth of the Mackenzie to the head of steamboatnavigation on the Athabaskaat Fort MeMurray. The height of land betweenit and the valley of the Saskatchewanto the south is but slightly over two thousandfeet in elevationand presentsan almost uniform flat surfacewith not even a ridge of hills to mark the changeof slopefrom the north to the south. The migratory birds of the Mackenzie Valley have the choiceof ttu'ee prinelpal routes as they return from their winter homes. They can come from the south, through Alberta, westernSaskat- chewan, western Montana and Utah, where Great Salt Lake, the winter homeof thousandsof birds,lies directly southof Great Slave Lake. A secondroute passesup the l•aeifie coast of the United States to Washingtonand thenceup the valley of the Columbiato the headwatersof the Athabaskaor up the valley of the Fraser to the watershedof the l•eaeeRiver. The third route is up the Miss- issippiRiver to southernor centralMinnesota; thenceto the valley of the Red River of the North and up the Assiniboineand Saskat- chewan Rivers to the sourcesof the Athabaska in Alberta, or across Saskatchewanat right anglesto the valleysof theserivers directly to Lake Athabaska. The first of these routes lies across a wilderness of mountains with many divides8,000 to 10,000feet high, and in the southern half of the United States through a district largely a desert. It would therefore seem probable that comparativelyfew species would employthis route and indeednot a singlespecies is known certainly to migrate from Arizonaor Utah to the MackenzieValley, and from a study of the data it seemsthat hardly half a dozen speciescan possiblytravel this route. There is a modificationof this route whleh ought apparentlyto form a convenientand fairly direct coursefrom Mexico or Texas to the MackenzieValley. This is alongthe foothills of the Rocky Fig. 1. Principal •iigra•ion ]•ou•es •o •he •i•ckenzie V•lley,

443 444 COOKE,Bird Migration in theMackenzie Valley. LOci;.['Auk

Mountainsthrough Colorado, Wyoming and Central Montana to Western Saskatchewan,over an almostuniform plain 5,000-6,000 feet in altitude. Probably somespecies do follow this route, but no positiveproof of thishas been found, and it is knownthat several speciesfor which this would seemto be the most desirablepath, actually go many hundredmiles out of their way to travel a route farther east. The birds that comeinto the Rocky Mountains are for the mostpart birdsthat are to breedthere or to go north only into the southernparts of Canada; very few go even as far north as northern Alberta. Hence in consideringthe probable routes of migrationto the MackenzieValley we canignore the usualnorth and south direction and consider that the bird comes either from the southeast or the southwest. The secondroute from Californiato the Mackenzieis the shortest of the three. No desertsor high mountainsintervene and the whole countryseems well adaptedto supporta wealthof bird life. If this route was largely used, then the birds of the Mackenzie Valley would be most closelyrelated to the speciesof the westernUnited States. Sincethe contrary is the fact, very few migratory western birdsoccurring in the MackenzieValley, it followsthat only a few speciescan use this secondroute. The third route is the longestand seemsquite roundaboutto one ;vhois in the habit of thinking of migration as always a north and south movement. In the springmost migratory birds enter the United Statesalong the north shoreof the Gulf of Mexico bem'een Florida and Texas. Of thesethe largerpart enter in a still smaller path, six hundredmiles wide, the middle of which is the mouth of the MississippiRiver. At the. border of the United Stat•, the courseof migration divides:part of the birds travel northeastto New England,the Gulf of St. Lawrence,and to Labrador'sinhos- pitable shores;a secondpart migratestraight north to the Great Lakesand HudsonBay; the third part moveat first north nearly tothe northern boundary ofthe United States and then turn north- west to the valley of the Mackenzieand evento Alaska. This last describedroute is the principalhighway for the migratorybirds of the MackenzieValley andis the mostnatural and notablemigration route on thewhole globe. Stretchingfor morethan three thousand milesfrom the mouthof theMississippi to the mouthof the Macken- Vol. 1915XXXII] j COO•:•,Bird Migration in theMackenzie Valley. 445

zie, not a mountainchain or evena ridgeof hills interfereswith the uniform movementsof the birds. The highestelevation is less than two thousandfeet, and so gradual are the slopesthat, with a few short portagesthe whole distancecan be traversedin a canoe. The wholeregion is well wateredand well timbered,affording ideal conditionsfor the supportof the multitudesof birds which swarm alongthis route as they do nowhereelse on the North American Continent. If the mouth of the Mackenzie River was due north of Louisiana and in the middle of the continent, bird migration by this route would be a uniform progressionfrom south to north in the spring and the reversein the fall. On the contrary the valley of the Mackenzie lies nearly two thousandmiles nearer to the Pacific than to the Atlantic Ocean,and the warm Japan current produces conditionsthat interferewith the uniformityof migrationan:t bring aboutvariations, probably not equalledanywhere else in the world, both in the directionand the speedof migration. That this diagonalnorthwest and southeastroute is traversed by birdsfrom the MississippiValley is shownpositively in the case of thirty-three species,for thesebreed in the MackenzieValley and passin migration acrossthe United States and yet occurin the United Statesas far west only as the easternedge of the plains. Henceit is certainthat thesethirty-three species have a northward migrationin the MississippiValley from easternKansas to western Minnesota and thence a northwestward route to Lake Athabaska. This is shownon the accompanying•nap of the distributionof the Rose-breastedGrosbeak (Zamelodia ludoviciana). It is evident that the westernmostbreeding birds, those that summerin the MackenzieValley must havereached their breeding groundsfrom the southeastby way of the MississippiValley. The causeof the choiceof this route is easklyfound in the con- ditionsof moistureand woodland. All thesespecies are either lovers of damp forestsor of moist'meadows and marshy lakes. Their favoritesurroundings extend in the United Statesnot farther west than eastern Kansas and western Minnesota. On arriving at Manitoba, the dry plainsthat have beena barrier on their left for the last thousandmiles, become better watered and interspersed with grovesand soonthese groves unite to form almostcontinuous ! ,

• Breed•n9 rc•n9e. .... WesternDor•Jer o•re•qulmr r•nBe j• mi9ro%ion

Fig. 2. Rose-breastedGrosbea• (Zamelodialudo•:iciana). 446 Vol. 1915XXXlI] j COOKE,Bird Migration in the Mackenzie Valley. 447 well-watered forest--a genuine birds' paradise. Attracted by the early seasonand abundantfood supply,the birds turn m)rth- westwardand settle for the summerin the valley of the Mackenzie. The thirty-three speciesthat traverse this route are Sterna hitundo, Micropalamahimar•topus, Limosa hwmastica,Numenius borealis,B•ztco b. borealis,Sphyrapicus v. varius,Chordeiles v. virginl- a;,us,Sayornis phoebe, Empidonax fiatgvcntris, Empidonax trailli alnorum, Otocorisalpestris hoyti, Cyanocittac. cristata,Euphagus caroli•,v.s,Carpodacus p. purpurcus,Calcarius 1. lapponicus,Passer- herbulusn. nelsoni,Zonotrlchla albicollis, Spizella m. monticola, Mclospizam. mclodia,3ielospiza georgiana, Passerella i. iliaca, Zamelodlal•.dor'iciana, Vireosylva philadclphica, Lanivireo s. soli- tari.us,Dendroica tigrina, Dendroicamagnolia, Dendroica castanea, Dendroica virens, Dendroica p. palmarum, Seiurus aurocapillus, IVilsonia p. pusilla, Hylocichlaa. aliciw, and ttylocichlaguttara pallasi. A modificationof this route from the southeast in • stillmore pronouncedform is followedby the Bqlite-wingedScorer (Oidemla dcglandi),which winters off the coastfrom Massachusettsto New Jerseyand in its springmigration follows the valleyof the Connecti- cut, crossesto the Hudson, thence to the Great Lakes and north- westward to its summer home in the Mackenzie Valley. It is probablethat thissame general route is followed by manythousands of the ducksof otherspecies whleh winter so abundantlyalong the coast from ChesapeakeBay to Florida, but it is also trne that a comparativelysmall part of thesetraverse this route asfar asGreat SlaveLake, sincethe larger part stopfor the summerin the "ducks' paradise" of Manitoba and the Saskatchewan. Still anotherroute, practicallyeast and westinstead of north and southis followedby the threespecies of Jaeger,which winter on the Atlantic, appear in Hudson Bay with the earliest open water and then crossnearly due west to the breedinggrounds about Great Slave Lake, and to the northward. There are eighteenspecies that alsoprobably use the main route from the southeastbut the proofof this useis not sosimple, because thesespeeies not only occurin the central MississippiValley, but alsorange regularly across the plainsto the foothillso• the Rocky Mountains. These eighteenspecies are: Grus americana,Pisobia 448 Coo•, BirdMigration in theMackenzie Valley. [Auk[oct. fuscicollis,Ereunetes pusillus, Bartramia longicauda,Falco s. spar- verius, Colaptesauratus luteus, Tyrannus tyrannus, Empidonax miminus, Agelaius phozniceusfortis, (2uiscalusquiscula a'•eus, l•ircosylvaolivacea, Mniot,ilta varia, Vermi•,orac. celata,Vermivora pcregrina,Dendroica striata, Setophaga ruticilla, ttylocichla ustulata swainsoni,and Planesticusm. migratorius.

Fig. 3. Red-eyed Vireo (Vireosylva olivacea).

They presentsome of the mostinteresting problems in the study of bird migration. In the caseof the Red-eyed Vireo, the species breedsalong the whole northern tier of stateswest to Washington, ¾ol.1915 XXXII ]1 COOKIe,Bird Migration in theMackenzie Valley. 449 and occursin migration in Colorado and Wyoming, but the indi- viduals that appear first in Alberta are not birds that have passed through Colorado,Wyoming and Montana, becausethe dates of migration, as indicated on the map, show that migration is early and rapid in the middle MississippiValley and late and slow along the foothillsof the Rocky Mountains. Th• earliestarrivals in southeasternSaskatchewan appear on the average May 17, but they do not comefrom directly south, sincein easternColorado, 600 miles to the southwardthe birds do not appear on the average until May 22. The first advanceto AthabaseaLake by May 28, which is just the date at whi :h they appearon the averagein north- ern Montana and northern Idaho. Therefbre it is evident that the birds of the Mackenzie Valley at Lake Athabaska have not comeby way of the Rocky Mountains, but by the route near the MississippiR•ver. The samegeneral method of proof can be applied to the migra- tions of the other seventeenspecies in this group. The proof is particularly clear and convincingin the caseof the Robin, Flicker, Bronzed Grackle, Redstart and the Black-poll, Tennesseeand Black-and-White Warblers. Eighty-twoother species of migratorybirds breed in the Macken- .zieValley and during the winter or in migrationoccur across the United States from the Atlantic to the Pacific. But even with thesespecies it can be shownthat mostof them probablyreach the MackenzieValley from the middle part of the MississippiValley. These eighty-twomay be divided into fifty-eight speciesof wide ranging water birds, nine speciesof hawks and owls, and fifteen speciesof smaller land birds. The fifty-eightspecies of waterbirds that arefound at Lake Atha- baskaand whichalso range from the Atlantic to the Pacificoceans presenta problemwith regardto their route of migrationthat with the recordsat handcannot be certainlysolved. The CanadaGoose is one of the abundant water birds of the Mackenzie Valley and may be takenas representative of the abovementioned group. The great bulk of the MackenzieValley CanadaGeese must come from the MississippiValley, where the speciesis abundant,for Branla canadensisis rather rare on the Pacificcoast, its placethere and in Alaska being taken largely by the other three forms, htttchinsi, occidentalis and minima. 450 COOK•,Bird Migration in theMackenzie Valley. tOct.[Auk

But the fact that the MississippiValley birdspass to the Macken- zie Valley cannot be proved by dates of migration as is shownby the followingrecords. The CanadaGoose begins its migrationnear the •YIississippiriver in February. .Aboutthe twentieth of that month may be consideredthe date of its normal arrival in south- easternIowa. Passingslo•vly north it appearsone month later in southeastern•innesota at the end of the first week in April. •i- gration in southeasternNebraska commencesat about the same time as in southeastern Iowa, but the birds move north a trifle fasterand crossto Saskatchewanabout the first of April. Further west the Canada Goose winters not far south of the United States boundaryand crossesinto southernAlberta the last of March. On the Pacific coastthe specieswinters in British Columbia. When, therefore,it is known that the Canada Goosearrives at Lake Atha- baskaApril 20, no certainconclusion can be drawn from this data as to whether these earliest birds come from Manitoba, Alberta, or British Columbia. The last furnishes its most northern winter' home and hencewould require the least rapid migration in spring to reachLake Athabaskaby the givendate. The journeyby way of Manitoba or easternSaskatchewan is a longerdistance and the. later start requiresa higherspeed of migration. Henceif no other informationwas available,the migrationdates alone would lead oneto supposethat the earliestbirds at Lake Athabaskacame from the southwest. But as stated at the outset, the relative numbers, of the birds east and west of the Rocky Mountains, make it certain that most the birds of Lake Athabaskareally do comefrom the •VIississippiValley. Since this is true of the geese,it may be assumed to be true also of the Mallard and Pintail Ducks which travel in companywith the geeseand have the samerange from the Atlantic to the Pacific. The samemethod of reasoningmay be appliedto the largerpart of the fifty-eightspecies of wide-rangingwaterfowl that occurregu- larly at Lake Athabaska. Most of them are abundantin migration acrossthe moistplains from Kansasto Saskatchewan,but are com- parativelyrare in the wholemountainous region of westernUnited Stateswhere favorable localities either for breedingor for feeding duringmigration are few and of smallarea. Henceit mustbe true in generalthat the untold thousandsof water birds that frequent VoL 1915XXXII] j COOKS,Bird Migration in theMackenzie Valley. 451 the lakes and marshesof the Mackenzie Valley, come from the moisterportions of the MississippiValley. With regardto the nine speciesof migratoryhawks and owlsof wide range that visit Lake Athabaska nothing can be judged at present either from distributionor migration as to the route or routesthey employ. Fifteenspecies of the smallermigratory land birdshave the samerange and in somecases, the datesof migration afford a hint of the route traveled. Thus in the caseof the Crow, the dates of migration showclearly that the earliestindividuals to reach southwestern Manitoba come not from South Dakota as would be expected,but from the timbered regionsof Minnesota. The date of arrival at latitude fifty degreesin southwesternMani- toba is March 27, as determinedby fourteen years' observations from four neighboringtowns. March 27 is a fair averagedate for the arrival of the Crow in east central South Dakota, three hundred and fifty milesto the southward;while the averagedate of arrival in southern North Dakota is a week later than in Manitoba. Con- tinuingin the samegeneral northwestern course it is probablethat the Crows appearingApril 2, 1893, at Osier, Saskatchewancame from Manitoba rather than from Montana; since this date would be consideredan early date of arrival in southeasternMontana. In this mannerthe datesof migrationshow that the earliestCrows in the Mackenzie Valley come from the wooded districts of the MississippiValley. The Myrtle Warblerpresents a quite similarset of dates. This speciesranges from the MississippiValley throughoutthe Rocky Mountains and to the Pacific, becomingmuch rarer west of the mountains. In its springmigration, it reachessouthern Manitoba April 23, at about the sametime as its first appearancein central Nebraska and northern Colorado,showing conclusively that the Manitoba birds comefrom the southeast. It is equally sure that the arrival at Osier, Sask.latitude fifty degreesMay 4, 1893, came from the southeast,for this is the usual time of arrival in central Montana latitude forty-sevendegrees. Farther north a new possi- bility presentsitself, sincethe May 16 birds of Lake Athabaska might, as far as the date is concerned,have come from British 'Columbiain the southernportion of which Province they arrive the middle of April. To determinethis latter point use can be 452 Coox•,Bird Migration in theMackenzie Valley. rock.[Auk made of the generalprinciple of parallel lines of migration. Since it is true that the birds of westernMinnesota passnorthwestward to Saskatchewan,it is probable that the birds of Montana also proceedin a northwesterndirection and traverseAlberta, in which easeit is altogetherunlikely that the birdsof the samespecies would be migrating northeastfrom British Columbiato Athabaskaacross the route of the easternbirds and at right anglesto it. A more reasonableassumption is that the birdsof BritishColumbia migrate also northwestwardand proceedto Alaska, where the dates of arrival, May 5 at Fort Relianceon the middle Yukon and May 18 at the mouth of that river, show that the birds must have come from British Columbia. The migrationrecords now availableare insufficientto determine whether the remainingthirteen speciesof small migratory wide- rangingland birds come to the Mackenzie Valley from the south- east, or southwest. There remain twenty-three speciesof migratory birds which breedin the MackenzieValley, but whichin migrationare confined to the Western United States, ranging not farther east than the easternedge of the plains. Thesemay be dividedinto three groups. The first groupcomprises nine speciesthat in winter are confined to the Pacificcoast; the secondgroup, three speciesthat rangeeast to the Rocky Mountains, and the third, sixteenspecies that occur east to the plains. The speciesof the first group,whose routes are best known, are the Pacific Eider (Somateriav-nigra), Black Brant (Bravuranigri- cans), Short-billed Gull (Larus brachyrhynchus),Ross's Goose (Chen rossi), and the Northern Varied Thrush (Ixoreus ncevius meruloides). The aeeompanylngchart showsthe principal migra- tion route followedby each of these species. They all crossthe Rocky Mountains, but in widely separatedlatitudes. The Ross's Goosecrosses the lofty rangesof the main chainof the RockyMoun- tains from northeastern California to northwestern Montana and thencenorth acrossthe MackenzieValley to its breedinggrounds on the Arctic islands. The Northern Varied Thrush winters mainly in the interior of Californiaand in westernNevada. Its main migrationroute crosses thence through northern Idaho and northwesternMontana and I I I I

i

Fig. 4. MigrationRoutes from the PacificCoast to the MackenzieValley. 453 4:54 COOK•,Bird Migration in the Mackenzie Valley. loc•.[Auk along the mountainsof easternBritish Columbiato the valley of the Liard and reachesthe Mackenziea few milesbelow Fort Simp- son. The relation of the breedingand wintering areas of the Short- billed Gull makesit probablethat the main migrationroute follows up the Fraser River and down the Peace River to the breeding groundson Great Slave Lake and northware]. The Black Brant is a common breeder at the mouth of the Mac- kenzie and along the coast to the eastware]. It does not reach •lackenzie from the south, indeed it is not known inland in that Province, but in springmigration it passesup the Pacific coast to the mouth of the Yukon, up this river to its junction with the Por- cupine,and up this stream and acrossthe low divide to the mouth of the Mackenzie. It seemsprobable that the Pacific Loon (Gayla pacifica)and the SabineGull (Xema sabini)follow this sameroute, but the proof is not as yet conclusive. The Pacific Eider (SomateriaV-nigra) doesnot occurinland in either :Alaska or Mackenzie. It winters around the Aleutian Islands and is a common breeder on the coast east of the mouth of the Mackenzie. Hence it followsthat the line of migration must pass through Bering Strait and go round the northern coast of Alaska. It seemscertain that the King Eiders (Somateriaspecta- bilis) and the GlaucousGulls (Larus hyperboreus)breeding on the Mackenzie coast arrive by the same route rather than from the Ariantie side. Indeed it can be said there is nothing to indicate that any birds migrate regularly from the Labrador coast north- westerlyto the coastof westernMackenzie. A migrationroute as yet unsolvedis that of theYellow-billed Loon (Gayla adamsi). It appearsat Great Slave Lake as soonas any part of the lake is open. It is not knownat any time of the year, either east, south, or west of Great Slave Lake, and at the time it appearsthere, no openwater existsanywhere between that Lake and the :ArcticOcean, and the speciesis n6t yet recordedfrom any- where inland in Alaska. In fact the records as they stand at presentare explainableonly onthe theoryof a singleflight from the openPolar Sea to the summerhome on Great Slave Lake, and sucha flight is scarcelybelievable. The three western speciesoccurring in the Mackenzie Valley Vol. 1915XXXII] j CooKs,Bird Migration in the Mackenzie Valley. 455 that range from the Rocky Nfountainsto the Pacific are Spizella passerinaarizonce, Vireosylva gilva swainsoni, and Pirangaludovici- ana. The migrationroute of this last speciesis especiallyinterest- ing. It breedsover the wholeof westernUnited Statesfrom the eastern foothills of the Rocky Mountains to the Pacific. Hence if one saw the map of the breedingrange and noted that the line of the easternmost limit was almost north and south and extended without a break from Nfexico to Canada, he would take it for granted that the breeding birds of Alberta reached their summer homeby a migration route along the eastern slopeof the Rocky Mountains. Suchreasoning is correctwith almostall species,but an examinationof the large amount of migration data available showsthat the WesternTunaget is an exceptionto the rule. The bird wintersin Guatemalaand when it starts north in the spring the individualsalong the Pacific coastmove north faster than thosethat chooseto migrate alongthe easternslope of the Rocky Nfountainsas shownby the isochronallines on the accompanying map. By Nfay 10, the earliest migrants have reachednorthern Washingtonalong the Pacific coast,while in the Rocky Mountains they are just entering southernColorado. During the next ten days the eastern birds loiter across Colorado to southeastern Wyoming,while on the samedate, Nfay 20, the first birds appear in central Alberta, a thousand miles farther north. It is evident that these latter birds could not have comeby way of Colorado, but must have come from Washington and British Columbia, though this latter assumptionrequires that they crossthe main chain of the Rocky Nfountainsat a time in the spring when even the lowestpasses are still coveredwith snow. It is true that warm weather has already come by this date in the southern Mackenzie Valley, but it is one of the strangestproblems in bird migration as to how the Western Tanagersknow that on the other side of thosesnow clad rangessummer is waiting for them. The migrationdates of the WesternChipping Sparrow show that the breedingbirds of Alberta followthe samegeneral route as out- lined abovefor the WesternTanager, while the data sofar available concerningthe migrationof the WesternWarbling Vireo throw no light as to the route employed. The fifteenwestern species breeding in the MackenzieValley and Fig. 5. Migration Route of the Western T•n•ger (Piranga ludoviciana). 456 Vol. 191.5XXXII] I CoorrE,Bird Migration inthe Mackenzie Valley. 457 rangingto the easternedge of the plainsin the United Statesare: Larus californicus,Branta canadensishutchinsi, Grus canadensis, Macrorhamphusgriseus scolopaceus, Sayornis sayus, Xanthoccphalus xanthocephalus,Calcarius lapponicusalascensis, Calcarius pictus, Poo•cetesgramineus confi•is, Passcrculussandwichensis alaudi•us, Passerherbuluslecontci, Zonotrichia Icucophrysgarbbell, Spizclla mo•ticola ochracca,Spizella pallida, and Seiurus noveboraccvsis notabilis. Larus californicuswinters abundantly on the coast of British Columbiaand breedscommonly from Great SlaveLake northward, showingthat its route of migration is northeast acrossBritish Columbia. Branta ca•mdensistmtchinsi is known to migrate in immense flocksfrom the plainsof the Mississippito thoseof the Mackenzie on its way to its northwesternbreeding grounds. The sameis true of Macrorhamphusscolopaceus, the line of whosenorthwestern mi- gration is known to extendfrom Florida to Great Slave Lake; the sameroute is undoubtedlyfollowed by Gruscanadensis. The earliestindividuals of Sayornissayus reach southern British Columbia about two weeks earlier than the first arrive in southern Colorado. The Alberta datesagree with thoseof Montana rather than those of British Columbia and Washington,while the dates on the lower Mackenzie and the Yukon can be satisfactorilyex- plained only on the suppositionthat these birds have come from British Columbia. Xanthocephalusxanthocephalus is so rare at Athabaska Lake and northward, that it can be consideredas hardly morethan a stragglerin the MackenzieValley. The date of arrival ß in eastern Saskatchewan however is so much earlier than in eastern Montana, as to indicate that the birds of Saskatchewancome from the southeast. The migration records of Zonotrichialeucophrys gambeliare so much earlier in British Columbia and Alaska than in correspondinglatitudes to the eastwardas to makeit practically certain that the Alaska birds have come by way of the interior warm valleysof British Columbia. Hence it is equally probable that the individualswhich swarmin Nebraska,Kansas, and south- ward during migration and winter kre the birds that pass north through Manitoba and Saskatchewanto breed in the Mackenzie Valley. [Auk 458 CookE, Bird Migrationin theMackenzie Valley. [Oct.

The migrationroute of Calcariuspictus is evident sincethe great majority of the individualsare confinedduring migration and winter to a narrowbelt of plainscountry extending from Texasnorthward to the Saskatchewanand thencenorthwestward along the Macken- zie to the summer home on the Barren Grounds of the Arctic. The same route is followed more or less dosely by Passerherbulus lecontei,Spizella pallida, and Calcariuslapponicus alascensis. One of the strangestmigration routes to the MackenzieValley is that of the Yellow-belliedSapsucker (Sphyrapicus v. varius). The mouth of the Nahanni River a hundred miles northwest of Fort Simpsonforms the normal limit of the species'range in that direc- tion; this is in longitude124 ø W. The Yellow-belliedSapsuckers breedinghere comeby way of Minnesotaat least as far east as longitude96 ø W. The rangeof this speciesextends thence south throughwestern Missouri on the meridianof 94ø and then south- westthrough the wholeeastern half of Texas to at least longitude 103ø W. in southwesternMexico. Thus the migrationroute forms a bow,the southernhalf extendingthrough nine degrees of longitude in a portionof the globewhere this is equalto six hundredmiles, and the northernhalf through thirty degreesof longitude,equi- valent in thosehigh latitudesto nine hundredmiles. The above routes accountfor the different groupsof species breedingin the MackenzieValley as follows: Easternspecies ranging in the centralUnited Statesonly to the edgeof the Plains. 33 Easternspecies ranging regularly or occasionallyto the RockyMountains in the centralUnited States. 19 Speciesthat in migrationor duringthe winter occuracross the whole United States from the Atlantle to the Pacific. 82 Western speciesconfined almost entirely to the Pacific coast. 9 Western species,ranging only occasionallyeast of the RockyMountains in the centralUnited States. 3 Westernspecies, ranging to the easternedge of the Plains in the central United States. 15 Total migratoryspecies breeding in the MackenzieValley, and wintering to the southward. 161 In addition to these, the list of the birds of the Mackenzie Vol. 1915XXXII] j RATnBV•,Wares Birds of thePuget Sound. 459

Valley includesseveral groups of speciesthat do not come under any of the aboveheadings. Non-migratory speciesand thosewhich occur during the winter. 37 Speciesthat barely reachthe MackenzieValley from the south,being found on the Athabaskaand not rangingnorth to Lake Athabaska. 41 Speciesnot includedunder the previousheadings, being for the most part stragglersor speciesthat are most common on the Arctic islands. 29 Total speciesknown to occurin the MackenzieValley 268 The distinctlyeastern character of the avifaunaof the Mackenzie Valley is shownby the fact that of the onehundred and sixty-one regularbreeding migrants, only eighteen are known with certaintyto reachthe Valley from the westor southwest,while it is knownwith equalcertainty that seventy-onereach it from the southeastor east; and of the remaining seventy-twospecies, mostly water-birds, probablyfour-fifths come from the centralMississippi Valley.

LIST OF WATER AND SHORE BIRDS OF THE PUGET SOUND REGION IN THE VICINITY OF SEATTLE.

BY SAMUEL F. RATHBUN.

Tins regionis a ,nuch favoredresort of ,nany of the speciesof water birdswhose habitat is the Pacificcoast and this undoubtedly is accountedfor by the fact that within its boundariesare e•n- braced the essential desiderata necessaryto attract them, viz. protectionfrmn the elements,an abundanceof food and a •nost equablete•nperature throughout the year. Its geographicallocation is likewisefortuitous being nearly in the direct line of migration of the countlessnmnbers of birds whose sramnet home is the North Pacific and of these, thousands use this regionas a winter resort,finding here every requirement