A REVISION OF THE FORMS OF THE GREAT BLUE HERON(ARDEA HERODIAS LINNiEUS).
By Harry C^. Oberholser,Of the Biological Survey, United States Department of Agriculture.
Large birds, as a rule, do not seem to prove generally attractive tothe modern systematic reviser, chiefly because of the usual difficultyof gathering together a satisfactory series, and of handling the mate-rial in making comparisons. The great blue heron, Ardea herodiasLinmeus, like many other large species of North American birds, hasbeen in much need of careful systematic study, and the results m thefollowing })ages are ofl'ered as an attempt at such revision. Previousliterature bearing on the subject is confined to descriptions of newforms, and to other short ]:)apers. Like the writer's recent review ofButondes virescens and its races,^ the present paper was undertakenat the suggestion of, Mr. H. W. Henshaw, chief of the Biological Sur-vey, as an aid in working out the ranges of the forms of this species fora forthcoming bulletui by the Biological Survey on the distributionand migration of the North American herons.Notwithstanding the difficulty of obtaining material, it has beenpossible to examine altogether 221 specimens of Ardea herodias,including the types of all its valid races except t:wo, Ardea herodiasherodias and Ardea. herodias lessonii. This material comprises theentire collections of the United States National Museum, with that ofthe Biological Survey; the American Museum of Natural History;the Academy of Natural Sciences of Philadelphia, with those ofMr. Witmer Stone and AIi*. H. W. Fowler; the collection of Dr. Jona-than Dwaght, jr.; also a number of specimens from the Museum ofComparative Zoology at Cambridge, Massachusetts; from Messrs.A. E. and O. Bangs; the Field Museum, of Chicago; Mr. John E.Thayer; and Mr. Edward A. Preble. In addition, Mr. Joseph Grin-nell very Idndly made some comparisons of specimens in the Museumof Vertebrate Zoology of the University of California, also sent meas-urements of a number of specimens; and Mr. G. Willett has furnishedinformation concerning the occurrence of the species on the Santa
' Proc. U. S. Nat. Mus., vol. 42, 1912, pp. 529-576.Proceedings U. S. National Museum, Vol. 43?No. 1939. 631
532 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.43.Barbara Islands, California. To all of these institutions and indi-viduals the writer wishes to express his obligation; also particularlyto Dr. Charles W. Richmond, the Assistant Curator of Birds in theUnited States National Museum.The geographical range of Ardea herodias, as a species, extendsfrom southern Alaska and southern Canada, south through theUnited States, Mexico, Central America, and the West Indies, tonorthern South America, Although the available specimens fromsome regions are few, there are apparently 10 recognizable races,which subsequent investigation of abundant material may increaseby one or two. There are already six current forms, so that the pres-ent investigation has resulted in the addition of four. The greatblue heron lives in almost any sort of country, forest or open, desertor humid, if it has only the one requisite?water, from which itobtains the major portion of its food. It breeds usually in colonies,in trees or on the ground, and doubtless wanders far in search offood, along the shores, shallows, and muddy banks of streams,lagoons, and ponds, which are its favorite hunting grounds. Toolittle is known of the details of distribution to permit a very decidedopinion on the life zone affinities of the various subspecies, but fromwhat we Icnow it does not appear that, except in a general way,theyconform very well to accepted zonal boundaries, as suitable nestingsites are often the controlling influence. Conseq.uently, the attemptto give them some such status must be taken with proper reservation.Most of the races are more or less migratory, though some of thismovement is doubtless the well-known presestival and postsestivalwandering in which herons so commonly indulge. Two forms, Ardeaherodias cognata, and Ardea herodias oligista, ^ with probably also Ardeaherodias sanctilucae, and possibly ^4 reject herodiasfannini, are sedentary,or at least do not pass beyond their ascribed breeding areas. As in somany other wide-ranging species, some of the far-separated subspeciesresemble each other much more than they do the adjacent forms withwhich they must bear nmch closer phylogentic relationship. Forinstance, the West Indian race, Ardea herodias adoxa,^ is much morenearly like Ardea herodias herodias than like the intervening Ardeaherodias wardi; Ardea herodias cognata, from the Galapagos Islands,very much more closely resembles Ardea herodias treganzai, from thesouthwestern United States, than it does Ardea herodias lessonii ^from Mexico; Ardea herodias oligista ^ and Ardea herodias hyperonca *are both much more like Ardea herodias herodias from the easternUnited States than like the interposed Ardea herodias treganzai;while the Mexican Ardea herodias lessonii^ is closest in appearanceto Ardea herodias fannini of British Columbia, instead of to the
? See p. 553. 2 See p. 544. ? See p. 555. * See p. 550.
NO. 1939. REVISION OF ARDEA HERODIAS?OBERnOLSER. 533adjoining Ardea Tierodias treganzai, of Arizona, or the also interveningArdea herodias hyperonca ^ of California. All but three of the races,Ardea herodias sanctilucae, Ardea herodias cognata, and Ardea herodiasoligista,^ have a range of considerable, in some cases, of wide extent,as seems to be often the case with large bu"ds which wander aboutover a large area. The size of the various races does not seemto correspond with any regularity to geographic conditions, forwhile the southernmost form, Ardea herodias cognata, of the Gala-pagos Islands, is of small size, the smallest race is Ardea herodiasoligista,- of the Santa Barbara Islands; whde on the adjacent Cali-fornia mainland is one of the largest, Ardea herodias hyperonca,^ be-yond which, in the northernmost subspecies, Ardea herodias fannini,lives a l)ird of only medium size. Furthermore, the FloridianArdea herodias wardi is larger than either Ardea herodias adoxa ^from the West Indies, or Ardea herodias herodias from the north-eastern United States,The adult female of Ardea herodias is identical in color with themale, but averages smaller. Individual variation is slight, and con-sists principally in size, and in the color of neck and back. Seasonalchange in plumage is likewise insignificant, probably because of thelittle wear to which the plumage is subject; and, though the necksometimes becomes a little lighter in late summer, the back andother parts seemingly undergo little or no change. Apparentlysoon after the j^oung are hatched, however, the color of the bareskin around the eye in the adult changes to yellowish green, andthe maxilla becomes almost entirely dusky olive, leaving only itstomia and the mandible yellowish.The young bird in first complete (i. e., juvenal) plumage differsconsiderably from the adult in having the whole pileum and uppersides of head dull, dark brown or brownish black; the entire neckand lower half of the sides of head much spotted or washed with buff",ochraceous, tawny, or chestnut; ground color of neck dull gray with-out any drab tinge; upper parts of body, wings, and tail, duller,more brownish, the long, plumelike development of dorsal andscapular feathers lacking; superior wing-coverts tipped or terminallyspotted wdth ochraceous or tawny; edge of wing paler and moremixed with white; thighs somewhat lighter; breast and abdomenwhite, streaked broadly with gray and slate color, and more or lessmixed with cinnamon rufous or buff; jugular plumes wanting; andthe tufts on sides of breast gi'ay striped with white; iris gambogeyellow; eyelids and bare horizontal lore space light apple green;maxilla black or blackish, the tomium brownish or horn color;mandible pale pea green, passing into clear horn 3'ellow on the ter-
i See p. 550. 6 See p. 553. * See p. 544.
534 PROCEEDiyOS of the XATIOyAL MUSEUM. VOL.43.minal half; tibia and soles of toes apple green; remainder of legs andfeet black. From this stage the bird gradually passes into the adultplumage, and specimens with all sorts of combinations of adult andyoung plumages are to be found. Some individuals in juvenalplumage are smaller than adults, but the young bird becomes fullygrowTi before assuming the adult livery.The diagnostic characters most useful ui separating the variousraces of Ardea Tieroiias are size, together with the shade of color onthe neck and upper parts. The posterior lower surface Ls practi-cally the same in all the forms. The juvenal plumage in most casesreflects the color differences of the adult, though usually to a lessdegree. All measurements are given in millimeters, and have beentaken as explained in the writer's paper on Butorides virescens} Inthe averages none but tvpical specimens have been used, wheneversuch could be obtained. Furthermore, in the tables of detailedmeasurements every specimen used in the diagnostic averages isindicated; and all other than adults are properly noted. The namesof colors employed are based on ^{i\ Robert Ridgway's Xomencla-ture of Colors. -In working out the races of Ardea herodias it became necessaryto determine as far as possible the status of Ardea occidentalis Audu-bon,^ Ardea wardi Ridgway,^ and the much discussed Ardea vmrde-mannii Baird,^ all from Florida, since this problem's solution bearsdirectly on the proper name for the race of Ardea lierodias inhabitingFlorida. The detailed evidence concerning these birds will bepresented in another paper.ARDEA HERODIAS HERODIAS Linnffius.[Ardm] herodias LrsN.EUs. Syst. Nat., ed. 10. vol. 1, 1758. p. 143 (America)(based on " Ardea fusca canadensis.'' Edwards [="The Ash-colour'd Heronfrom Xorth America. " Edwards], Nat. Hist. Birds, vol. 3. 1750, p. 135, pi. 135;vol. 4, 1751, p. 245 [Hudson Bay]; and, with a querj', on "Ardea cristatamaxima arnericana," Catesby. Nat. Hist. Carolina. Fla.. and Bahama Is.,vol. 2. Appendix. 1753, p. 10. pi. 10 [Virginia]).[Ardea] hudsonias Lixk^us, Syst. Xat., ed. 12, vol. 1, 1766, p. 238 (Hudson Bay)(based on "Ardea americae septentrionalis" [="The Ash-colour'd Heron fromNorth America"], Edwards, Nat. Hist. Birds, vol. 3, 1750, p. 135. pi. 135[Hudson Bay]; and on " Ardea freti hudsanis," Brisson, Omith., vol. 5. 1760,p. 407 [Hudson Bay]).Cfiars. suhsp.?Size rather small (compared with other races);neck and upper parts moderately dark in color.Description.?Adult male, Xo. 153329, U.S.X.M. ; Brewerton, XewYork, April 30, 1881; Edwin M. Hasbrouck. Sides of crown, and
1 Oberholser, Proc. U. S. Kat. Mns., vol. 42, 1912, p. 533.
- Ridgway, Nomenclature of Colors for Naturalists, Boston, 1S86.? Ardea occidental^ Audubon, Birds Amer. (folio), vol. 3, No. 57, 1835, pi. 281 (Florida Keys).< Ardea xcardi Ridgway, Bull. Nutt. Om. Club. vol. 7. January, 1S82. p. 5 (Oyster Bay, Florida).
? Ardea wurdemannii Baird, Rep. Expl. and Surv. R. R. Pac., vol. 9, 1858, p. 669 (southern Florida).
NO. 19;!9. REVISION OF ARDEA HER0D1A8?0BERH0LSER. 535long, slender, pointed, occipital crast black; forehead, center ofcrown, short vertical crest, superciliary stripe, malar region, chin, andmiddle of upper throat, white; sides of head and throat drab gray;neck all around deep drab gray, the middle line of foreneck streakedwith black, white, and cinnamon rufous ; upper surface, including taU,tertials, innermost secondaries, and superior secondary coverts, slategray, the long, narrow, plumaceous feathers of back and scapularspaler and glaucous, the median coverts and outer webs of greatercoverts paler gray, the rectrices somewhat brownish and distallydarker, inclining to slate color or blackish slate; primaries,secondaries (except the innermost), primary coverts, and alula,blackish slate, slate black or dull black; jugulum smoke gray, mediallywliite, streaked broadly with black, dull brownish slate, narrowlywith cinnamon rufous and pale cinnamon, the long, narrow, pointedj)lumelike feathers terminally pale drab gray, smoke gray or whitish;M tuft of black, partly white-striped feathers on each side of thebreast ; back of this a small patch of cinnamon rufous ; sides and flanksslate gray; breast and abdomen black, broadly streaked with white;lower tail-coverts white; thigLs and edge of wing cinnamon rufous,the latter shading to chestnut along the secondaries; lining of wingslate color, the axillars slate gray.In full spring plumage the iris is chrome yellow; the bare orbitalspace cobalt blue; bill wax yellow, brighter on the mandible, theculnien sometimes bro^vnish; legs black or dark brownish, the tibiausually more brownish ; soles of toes dull yellowish.Measurements.?Length (in flesh), 1067-1270 mm.; extent ofwing, 1790-1920 mm.; weight, 5-8 pounds.Male:^ Whig, 441-280 (average, 462.7) mm.; tail, 167-187 (176.6);exposed culmen, 123-151.5 (139.5); height of bill at base, 23.5-31.3(27.6); tarsus, 167-205 (183.6); middle toe, 93-114.5 (106.5).Female:- Wing, 433-471 (451.2) mm.; tail, 159-184 (173.7); ex-posed culmen, 127-146 (137); height of bill at base, 24.5-28.5 (26.7);tarsus, 157-194 (175.4); middle toe, 93-115 (102.4).Type-locality.?Hudson Bhj, Canada.OeograpJdcal distribution.?Eastern United States and southernCanada: In summer chiefl}^ the Upper Austral, Transition, and Cana-dian zones, north to Anticosti Island and Godbout, northeasternQuebec; Lake Temiskaming, central Quebec; Moose Factory andRat Portage, northern Ontario; Shoal Lake and Duck Mountain,central Manitoba; Osier, southern Saskatchewan; and Edmonton,central Alberta; west to southeastern British Columbia (probably);North Dakota; South Dakota; and probably eastern Nebraska;
' Ten specimens, from New York, Pennsylvania, New Jersey, Virginia, North Carolina, and Florida.s Twelve specimens, from New York, New Jersey, Conaeetieut, Maryland, South Carolina, Illinois, andTamaulipas.
536 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. 43.south to southern Alberta (probably); southern Saskatchewan(probably); eastern Nebraska (probably); central Iowa; centralIllinois; central Indiana; eastern Tennessee; and near Charleston,central eastern South Carolina; east to the Atlantic coast of theUnited States and of southern Canada, and to the Bermuda Islands.*Winters from Florida, northern Tamaulipas, and Texas, north to NewYork State; also on the Bermuda Islands. In migration or in winterit wanders west to Prospect Lake, British Columbia.The present subspecies appears to be very uniform over its entirerange, at least the specimens examined so indicate. There seems tobe little if any difference, in either size or color, between examplesfrom New York and Connecticut and those from North and SouthCarolina. A specimen from Castleberry, Alabama, taken November10, 1911, is large, but very dark, even for Ardea Jierodias Jierodias, andis nearer this than to Ardea Jierodias wardi. Another bird, taken atthe same place on November 18, 1911, is typical Ardea h. herodias.Two birds from Mount Carmel, Illinois (U.S.N.M. Nos. 84577 and84578), are slightly larger and slightly lighter above than typicalexamples of Ardea herodias herodias, inclining thus toward Ardeaherodias wardi, but they are much nearer the present form. As thesetwo were not breeding birds they were doubtless sojourners fromanother region, since the breeding form of this locality is Ardea herodiaswardi. The breeding bird of Saskatchewan and Alberta is probablyArdea herodias herodias, although no specimens are at hand. Animmature individual from Prospect Lake, British Columbia, Septem-ber 18, 1896 (No. 73573, Amer. Mus. Nat. Hist.), is just like immatureArdea herodias herodias, so far as I can see, though it may possibly bean intermediate between Ardea herodias treganzai and Ardea herodiasfannini. What is the breeding form of the Bermuda Islands is a littleuncertain, as I have examined no specimens from there, but theprobabilities favor its proper reference to Ardea herodias herodias.All specimens from Florida are, of course, only winter residents,and this State seems to be a regular winter home for members ofthis subspecies. It winters, however, with apparent regularityalso north to New York State, as there are several winter records ofspecimens from the States of New York, New Jersey, and Virginia.A single bird (No. 183328, U.S.N.M.) taken January 27, 1902, atCamargo, Tamaulipas, Mexico, on the Rio Grande, represents thesouthwestern limit of its known range. There is no certain recordof its occurrence in Louisiana, though it doubtless occurs there dur-ing winter. A bird taken September 17, 1886, by naturalists ofthe U. S. Fish Commission steamer Albatross, at sea about 130 milesoff the coast of New Jersey (lat. 38? 29' N. ; long. 71? 58' W.), showshow the species sometimes wandeis.
1 Probably this form.
NO. 1939. REVISION OF ARDEA HERODIAS?OBERHOLSER. 537The original description of Ardea Jierodias^ applies without doubtto the form of great blue heron inhabiting the northeastern part ofNorth America. Linnaeus' description was based on the "Ardeafusca canadensis,'' or "The Ash colour'd Heron from North America
"
of Edwards;^ and with a query on the "Ardea cristata maxiTnaamericana" of Catesby;^ both of which may be unquestionablyreferred to the great blue heron. Since Linnaeus gives only "Amer-ica" as the locality of his species, and since he apparently consideredthe identity of Catesby's bird doubtful, the type locality should,therefore, be Hudson Bay, from Edwards. The Ardea hudsonias ofLinnaeus * is a strict synonym of Ardea herodias, since it has in partthe same basis? ' ' Ardea americae septentrionalis " = " The Ash colour'
d
Heron from North America" of Edwards,^?and also the " Ardea fretihudsonis" of Brisson,^ both of which refer to the bird from HudsonBay.In all, 76 specimens of this form have been examined, from thefollowing localities, breeding records being indicated by an asterisk:British Colunibia.?Prospect Lake.Nova Scotia.?Newport.*Quehec.?Tadousac.Alahama.?CsisilebeTrj.Connecticut.?Liberty Hill; Rockville.District of ColumMa.?Washington.*Florida.?Gainesville; Oak Lodge (eastern peninsula oppositeMicco) (Feb. 1); Lake Harney (Dec. 3).Illinois.?Mount Carmel.Iowa.?Winnebago County.*Maine.?Island Falls; Brewer.Maryland.?Sandy Spring;* Glen Echo.*Massachusetts.?Dedliam.New Ham'pshire.?Hollis.New Jersey.?Haddonfield; Timber Creek;* Salem County;* Five
^lile Beach (Cape May County);* Cape May County; South Atlan-tic City; Atlantic City; Masonville; Mouth of Delaware Kiver (Jan.26); Atlantic Ocean off coast of New Jersey Gat. 38? 29' N.; long.71? 58' W.).New York.?Brewerton;* Springs (Dec. 20); Long Island (Dec.28); Shelter Island; Ardsley (Westchester County) ; Lawrence Coun-ty; Oakdale (Suffolk County); Consook Marsh, Highland Falls(Dec. 12); Rockaway (Dec. 14).
1 Linnseus, Syst. Nat., ed. 10, vol. 1, 1758, p. 143.2 Nat. Hist. Birds, vol. 3, 1750, p. 135, pi. 135; vol. 4, 1751, p. 245.? Nat. Hist. Carolina, Fla., and Bahama Is., vol. 2, Appendix, 1753, p. 10, pi. 10.* Syst. Nat., ed. 12, vol. 1, 1766, p. 238.6 Nat. Hist. Birds, vol. 3, 1750, p. 135, pi. 135.
.? Ornith., vol. 5, 1760, p. 407.
538 PROCEEDINGS OF THE NATIONAL MV8EVM. VOL. 43.North Carolina.?Currituck (Jan. 25); Hatteras.Pennsylvania.?Conneautee Lake (Erie County) ; Holmesburg (Phil-adelphia);* Bristol (Bucks County).*SoutJi Carolina.?Porcher's Bluff, Christchurch Parish *Virginia.?Cape Charles (Jan. 1); Strasburg; Dismal Swamp.*Wisconsin.?Waupaca.Tamaulifas.?Camargo.Measurements of specimens of Ardea herodias herodias.
Museum and No. Sex. Locality. Date. Collecttor. s
NO. 1939. REVISION OP ARDEA H^RODIAS?OBERHOLSER. 539
Measurements of specimens of Ardea herodias herodias?-Continued.
Museum and No.
Am. Mus. N. H.73573.
1
Am. Mus. N. H.45753.A.N.S.Phila.39659.
A.N.S.Phila.39660.A.N.S.Phila.2C.025.A.N.S.Phila.58251
A.N.S.Phila.349ir)A.N.S.Phila.48276.W.Stone 1207..Am. Mus. N. H.74173.U.S.N.M. 1.59980..U.S.N.MU.S.N .MA.N.S.Phila.26027
Sex.
Femalejuv.
Juv..
..do.
..do.
Juv.
Juv.
Juv.
Locality.
ProspectLake, Brit.Col.Long Island,N. Y.Mouth of Delaware Riv-er [N.J.?]doCape MayCo., N.J.South At-lantic City,N.J.M a sonville,N.J.Salem Co.,N.J.AtlanticCity, N.J.Cape Charles,Va.DismalSwamp, Va.Castleberry,Ala.doW i n nebagoCo, Iowa.
Date.
Sept.18, 1890
Jan. 26,1903
...doNov .27, 1878Sept.?, 1899
Feb. 6,1905Oct. 2,1892Jan. 1,1901June 20, 1897Nov.18,1911Nov.10,1911Aug.l9, 1879
Collector.
mm442
W.D.WinsordoAV. L. Ab-bott.C. W. Buv-ingerand J.A.G.Rehn.L.Rodgers..W.Hughes..
I. N. De Ha-ven.G.A.Smith.D. W. Pren-tiss, jr.W.MatthewsdoW. L. Ab-bott.
mm173
432 175
i430! 159432 164470 177
451445452445450464490435
180179174181177181184171
mm123
134.5133.5
mm27.5
25.525.5
272627.5
25.52527.528.527.528
mm161
159,5173
167170.5193
160176181179.519517929.819526.5 169
mm89.5
101.594.5111
103101106.510911710211499
ARDEA HERODIAS WARDI Ridgway.Ardea wardi Ridoway, Bull. Nutt. Orn. Club, vol. 7, January, 1882, p. 5 (OysterBay, Florida).Chars, snhsp.?Similar to Ardea herodias herodias, but decidedlylarger in all its measurements; upper surface and neck lighter in color.Measurements.?Total length (in flesh), 1295-1335 mm.; extentof wing, 1970-2090.Male:^ Wing, 486-518 (average, 497.7) mm.; tail, 181-209 (191.5);exposed culmen, 146-167 (156.9); height of bill at 'base, 28.5-32(30.8); tarsus, 195-232 (210.9); middle toe, 115-129 (119.9).Female: 3 Wing, 471-489 (477.2); tail, 173-192 (181.5); exposedculmen, 140-147 (143.7); height of bill at base, 27-30.5 (28.1);tarsus, 189-214 (205.4); middle toe, 111-123 (116.2).Type-locality.?Oyster Bay, Florida.Geographical distribution.?Southeastern United States to centralMexico : In summer chiefly the Lower Austral Zone, north to Hilton-head, southeastern South Carolina; southern Georgia; southernAlabama; Knox County, southwestern Indiana; Mount Carmel,southeastern Illinois; Henry County, southeastern Iowa; and Kansas;west to Kansas; central Oklahoma; Gainesville and probably Gurley,
' Full grown, but ia juvenile plumage.2 Fourteen specimens, from Florida, Georgia, Texas, and Jalisco.
^ Eight spexjimens, from Florida.
540 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.43.Texas; south to the Florida Keys and the Gulf coast of the southernUnited States, from Florida to Brownsville, Texas; east to the Atlanticcoast of South Carolina, Georgia, and Florida. Winters in Florida,southern Alabama, and Texas; also in Mexico, south to Ocotlan,Jalisco.This is a well-characterized form, and is easily distinguishable fromArdea herodias Jierodias. Even in the juvenal plumage it is usuallypaler on neck and upper surface than is Ardea h. herodias. A seriesof breeding birds from a cypress swamp near the mouth of the WliiteRiver, in the southwestern corner of Knox County, Indiana, and onespecimen from Mount Carmel, Illinois (No. 72837, U.S.N.M.), justacross the Wabash River, are smaller than Ardea herodias wardi fromFlorida, and in this seem to be rather closer to Ardea herodias herodias,yet in color they average much nearer Ardea herodias wardi. A singleexample from Henry County, southeastern Iowa (No. 12358, J.Dwight), and another from an unknown but probably easternlocality in Kansas (No. 72836, U.S.N.M.) are pale like Ardea herodiaswardi, and also large, too large, in fact, for Ardea herodias herodias.The present race thus doubtless occupies the southern IVIississippiValley, including eastern Kansas, eastern Oklahoma, and easternTexas. A breeding bird from Hiltonhead, southeastern SouthCarolina (No. 39031, U.S.N.M.), is like Ardea herodias wardi in color,but is smaller, and in this somewhat intermediate between Ardeaherodias wardi and Ardea herodias herodias. The same remarks willapply also to a specimen from Ossabaw Island, Georgia (No. 11902,J. Dwight), taken November 29, 1904. This Hiltonhead, SouthCarolina, record doubtless represents about the northernmost limitof the breeding range of this subspecies on the Atlantic coast, for thebreeding form about Charleston is Ardea herodias herodias. No speci-mens from Mississippi or from the central or northern portions ofGeorgia or Alabama Ifave been available, but, Ardea herodias wardiwithout doubt occupies in summer the southern part of all theseStates. Birds from Corpus Christi, Texas, seem to be identical inboth size and color with those from Florida. The following com-parative averages of millimeter measurements of specimens fromvarious parts of the range of Ardea herodias wardi show what differ-ences exist:
NO. 1039. REVISION OF ARDEA HER0DIA8?0BERH0LSER. 541
Localities. Wing. Tail. Exposedculmen. Heightof billat base. Tarsus. Middletoe.
Eleven males, from Florida and Georgia ,Two males, from TexasOne male, from Jalisco, MexicoSix males, from southwestern Indiana ,One male, from southeastern IowaOne male, from Kansas ,Seven females, from Florida ,Two females, from southwestern Indiana and south-eastern Illinois
mm499.1496486476486490477.2453
mm192.3185.5
.195182.2190184181.5185.5
mm158156.5146145.7153158143.7138
mm31.1302927.8292928.126.3
mm211214.5202183.4194202205.4180.5
mm119.8123115107.6119107116.2
This subspecies is evidently resident in Florida, as there is no WestIndian or Bahaman specimen ; and there are numerous winter recordsfor Florida. It winters also north at least to southern Alabama,probably farther in the Mississippi Valley. An example whichMessrs. E. W. Nelson and E. A. Goldman obtained at Ocotlan, Jalisco,Mexico, on December 25, 1902 (No. 184944, U.S.N.M.), seems to beundoubtedly of this form, and probably marks nearly or quite thesouthern limit of its winter range.It breeds rather early in the year, at least in Florida; commonly inMarch (Suwanee River, Lafayette County, March 21, 1890, Mr. F. M.Chapman; Fort Gardner, March 9, 1901, Dr. E. A. Mearns) ; and, occa-sionally, at least, by the middle of January, for Mr. C. A. Smith foundnest and eggs at Rutland, Florida, on January 15, 1890.The form of Ardea herodias here called Ardea herodias wardi wasfirst described by Mr. Ridgway as a full species,^ on the hypothesisthat Ardea Jierodias, Ardea occidentalis, and Ardea wardi should beregarded as three distinct species, of which the last two were consid-ered to possess both a white and a colored phase. That, however,Ardea wardi is but a subspecies of Ardea lierodias, is evident from theregular intergradation with Ardea lierodias lierodias that takes placewherever the ranges of the two closely approach, from South Carolinato Illinois. Furthermore, pure white birds {Ardea occidentalis) occuronly on the Florida Keys and the adjacent West Indian islands; andthe so-called Ardea wilrdemannii only where the ranges of Ardeaoccidentalis and Ardea wardi overlap. The complicated and puzzlingrelationships of these four birds
?
Ardea occidentalis, Ardea wiirde-mannii, Ardea wardi, and Ardea herodias?will be more fully dis-cussed in a separate paper. Suffice it here further to state that thewriter regards Ardea herodias as specifically distinct from Ardea occi-dentalis; Ardea wardi the Florida subspecies of Ardea herodias;^ Ardea
1 Ardea wardi Ridgway, Bull. Nutt. Orn. Club, vol. 7, January, 1882, p. 5.2 See Chapman, Bull. Amer. Mus. Nat. Hist., vol. 14, April 15, 1901, pp. 88-
542 PROCEEDINGS OF THE NATIONAL MUSEUM. VOL. 43.
occidentalis a distinct species; and Ardea vnlrdemannii a hybridbetween Ardea lierodias wardi and Ardea occidentalis.The writer has seen 60 specimens of Ardea Jierodias wardi fromthe subjoined locahties, an asterisk marking records of breedingbirds
:
Alabama.?Orange Beach, Perdido Ba}^, Baldwin County (Jan. 28).Florida.?Drigg's I^anding, Kissimmee River;* De Soto County(Dec. 8); Jacksonville; Lake Harney; eastern peninsula oppositeMicco; near Micco;* San Mateo;* Palatka (Jan. 29); Gainesville;Sebastian (Jan. 29); Lake Hatch-ne-haw;* Clive Key;* Fort Bas-singer; Punta Rassa (Mar. 10);* Suwanee River, Lafayette County;*Fort Gardner, Kissimmee River;* Clearwater; Seven Oaks; * BremerIsland, Lake Kissimmee;* Rutland; Tarpon Springs;* northernBrevard County; Amelia Island (Dec. 14); Amelia Island;* NewSm3a'na ; Hernando County.*Georgia.?St. Marys;* Mcintosh County ; * Ossabaw Island.Illinois.?Mount Carmel.*Indiana.?Cypress Swamp, near mouth of White River, south-western Knox County.*Iowa.?Henry County, 4 miles north of Hillsboro.Kansas.?(No definite locality specified.)South Carolina.?Hiltonhead.*Texas.?Corpus Christi.*Jalisco.?Ocotlan (Dec. 25).Measurements of specimens of Ardea herodias wardi.
NO. 1939. REVISION OF ARDEA HERODIAS?OBERHOLSER. 543Measurements of specimens of Ardea herodias wardi?Continued.
544 PROCEEDINGS OP THE NATIONAL MUSEUM. vol. 43.ARDEA HERODIAS ADOXA, new subspecies.Chars, suhsp.?Much like Ardea herodias herodias, but paler above,and apparently, at least in the female, of somewhat smaller size.Description.?Adult female, No. 97914, U.S.N.M.; Curasao Island,Caribbean Sea, February 10-18, 1884. Sides of crown, and occipitalcrest, black; forehead, center of crown, short vertical crest, super-ciliary stripe, sides of head, malar region, chin, and center of upperthroat, white; broad subauricular stripe pale drab gray; neck allaround deep drab gray, the middle line of foreneck streaked withblack, white, and cinnamon rufous; upper surface, including tail,tertials, innermost secondaries, and superior secondary ' coverts,rather light slate gray, the long narrow plumaceous feathers of backand scapulars paler and glaucous, the median coverts and outer websof greater coverts paler gray, the rectrices brownish on inner webs;primaries, secondaries (except the innermost), primary coverts, andalula, blackrsh slate, slate black, or dull black; jugulum smoke gray,medially white, streaked broadly with blackish slate, slate gray, andsmoke gray, the long, narrow, pointed, plume-like feathers terminallypale smoke gray or white; a tuft of black, partly white-stripedfeathers on each side of the breast ; back of this a patch of cinnamonrufous; sides and flanks slate gray; breast and abdomen, black,broadly streaked with white; lower tail-coverts white; thighs andedge of wing tawny chestnut, the latter shading to chestnut; underwing coverts partly slate color, partly slate gray, some with whitishshaft streaks; axillars slate gray.Measurements.?Probable male : ^ Wing, 460-468 (average, 464.7)mm.; tail, 177-185 (180); exposed culmen, 134-150 (143); height ofbill at base, 26.5-28.7 (27.7); tarsus, 178-187 (181.7); middle toe,109-113 (111.3).Female:^ Wing, 430-447( 440.5);- tail, 168-172 (170.5); exposedculmen, 121-131 (127.8); height of bill at base, 24-26,5 (24.8); tar-sus, 162-175 (167.5); middle toe, 92-108 (102).Type-locality.?Island of Curasao.Geographical distribution.?Bahama Islands, West Indies, Cura-sao, and Trinidad:^ resident in the upper Tropical Zone, north toAndros and Inagua islands, Bahama Islands; west to Cuba; south toJamaica; and east to Inagua Island, Bahama Islands. In winteralso north to Bimini and Berrj^ islands, Bahama Islands; east to NewProvidence Island, Bahama Islands, Santo Domingo, Porto Eico, andthe islands of St. Croix, Sombrero, St. Bartholomew, Barbuda, An-tigua, Montserrat, Guadeloupe, Dominica, Martinique, Barbados, St.
1 Three specimens, from the West Indian islands of Sombrero, Guadeloupe, and Grenada.2 Four specimens, from Inagua Island (Bahama Islands), Guadeloupe Island, and Jamaica, West Indies
' No specimens examined, but probably this form.
NO. 1939. REVISION OF ARDEA HER0DIA8?0BERH0L8ER. 545Vincent, Carriacou, and Grenada; south to the Isle of Pines, Curasao,and Trinidad.This race is of the size of Ardea herodias herodias or even smaller,and in color is somewhat intermediate between Ardea Jierodias herodiasand Ardea herodias wardi. It is thus decidedly smaller throughoutand rather darker above than Ardea herodias wardi. All the avail-able immature birds from the West Indies, excepting an adventi-tiously stained one from Jamaica, are fully as pale above as youngArdea h. wardi, and are even more different in color from the corre-sponding stage of Ardea herodias herodias than are adults. In sizethese immature birds are less than immature Ardea herodias herodias.Two breeding birds from Inagua Island, Bahama Islands, are appar-ently just like other West Indian birds.Specimens examined number eight, from the following localities inthe West Indies and Bahama Islands, breeding records being notedby an asterisk:Sombrero Island.Curasao Island.Jamaica.?Spanishtown.Grenada Island.?Telescope Swamp.Guadeloupe Island.Inagua Island (Bahama Islands).'^Measurements of specimens of Ardea herodias adoxa.
Museum and No. Sex. Locality. Date.
546 PROCEEDINGS OF THE XATIOXAL ilUSEUM. vol.43.Measwcments.?Total length (in flesh), 1100-11 70 mm.; extentof whig, 1755-1845.Male:^ Wing, 445-493 (average, 471.7) mm.; tail, 162-192 (179.9);exposed ciilmen, 132-157 (144.3); height of bill at base, 26-31 (28);tarsus, 165-193 (181.6); middle toe, 99-112.5 (105.9).Female: 2 Wing, 440-475 (455.5); tail, 164-182 (174.2); exposedculmen, 120.5-150 (137.2); height of bill at base, 25-28 (26.1); tarsus,157-183 (170.5); middle toe, 89-107.5 (100.7).Type locality.?Egg Island, Great Salt Lake, Utah.Geographical distribution.?Western United States to southwestern]\Iexico: In summer chiefly the Lower Austral, Upper Austral, andTransition zones, north to Douglas County, southern Wyoming;Grace, southern Idaho; probably also to Montana and eastern Wash-ington; west to central Washington (probably); central Oregon(probably); Nevada; Colorado River near Riverside Mountain, andPelican Island, Salton Sea, southeastern California; and the PacificOcean at the United States and Mexican boundary line;^ south toGardner's Laguna, Salton River, northern Lower California; Guay-mas, central western Sonora;* Santa Cruz River west of the Pata-gonia Mountains, Arizona; Fort Fillmore and Carlsbad, southernNew Mexico; Tornillo Creek, near Boquillas, central western Texas;and probably also northern Chihuahua; east to Kerr County, centralTexas (probably); Denver, central Colorado; Douglas County, east-ern Wyoming; and probably eastern Montana. Winters from Texasand Arizona south to Manzanillo, Colima, Mexico. In migrationwanders west to the Sacramento Valley, California, and east toCorpus Christi, Texas.This western race is apparently just like Ardea herodias ivardi incolor, but is decidedly smaller throughout. It differs from Ardeaherodias adoxa, of the West Indies, in somewhat paler upper partsand sHghtly larger size. It is, in fact, closer m characters to tliisthan to either Ardea herodias herodias or Ardea herodias wardi. TheJuvenal plumage of Ardea herodias treganzai is distinguishable fromthat of Ardea herodias herodias by its decidedly paler upper partsand usually paler neck.There seems to be no difference of consequence between birdsfrom various parts of this bird's range, as above defined. An imma-ture specimen from Bro^Misville, Texas (No. 30259, J. Dwight), isapparently this form; as is also a bird from Corpus Clmsti (No.79707, Amer. Mus. Nat. Hist.), taken at the late date of April 11(1889). We have only the head and part of the neck of an immature,
1 Fourteen specimens, from Arizona, New Mexico, Texas, Idaho, and California, including seven notexamined by the writer, but measured by Mr. Joseph Grinnell.2 Twenty-two specimens, from Arizona, Texas, Utah, Wyoming, Montana, Lower California, and Chi-huahua, including seven not examined by the writer, but measured by Mr. Joseph Grinnell.2 Probably not breeding here.
* Probably Ardea herodias treganzai; eggs in the United States National Museum.
NO. 1939. REVISION OF ARDEA HER0DIAS?0BERH0L8ER. 547apparently non-breeding example from the Pacific Ocean at theMexican and United States boundary hne (No. 133774, U.S.N.M.),taken July 16, 1894, but it seems to belong to this race, notwith-standing that the form from San Diego, Cahfornia, is not the same.No breeding examples from the San Joaquin Valley, or from thesouthern Sacramento Valley, California, have been available, and wetherefore can not determine the summer resident of this region. Asingle specimen without date, from the Sacramento Valley (No. 9472,U.S.N.M.), is apparently Ardea herodias treganzai, but it is doubtlessa migrant or a straggler.Altogether, 36 specimens have been seen, from the locaHties givenbelow, breeding records being followed by an asterisk:Arizona.?Sonoyta River at Quitobaquita (Feb. 4); Fort Verde;Gila River, 7 miles below San Carlos ; Santa Cruz River, west of Pata-gonia Mountains;* Fort Lowell; Tucson; San Bernardino Ranch;Colorado River at Monument 204 (Mexican boundary hne).*California.?Sacramento Valley; Pacific Ocean at the Mexicanboundary Une.IdaJio.?Grace.*Montana.?Gallatin Station.New Mexico.?huna; San Luis Springs.*Texas.?Fort Clark (Jan. 24); Corpus Christi; Brownsville (Feb.22).Wyoming.?Douglas.*CJiilmahua.?Rio Grande River near El Paso, Texas (Feb. 27) ; SanDiego.Colima.?Manzanillo.Louver California.?Gardner's Laguna, Salton River.*Sonora.?Caioii Bonito Creek, near the United States boundaryline. Measurements ofspecimens of Ardea herodias treganzai.
Museum and No. Locality. Date.
548 PROCEEDINGS OF THE NATIONAL MUSEUM. VOL. 43.Measurements
NO. 1939. REVISION OF ARDEA HER0DIA8?0BERII0L8ER. 549Type-localitij.?Espiritu Santo Island, Lower California, Mexico,Geograjjiliical distribution.?Upper Tropical Zone of southern LowerCalifornia, Mexico; south to San Jose del Cabo; north to San JoseIsland. Probably resident throughout the year.The present recently described form is easily separable from all theNorth American races, and seems most closely to resemble Ardeaherodias cognata Bangs, from the Galapagos Islands. It may bedistinguislied from Ardea herodias adoxa, of the West Indies, by itsmuch greater size throughout, and paler neck and upper parts; fromArdea herodias wardi by smaller size and more lightly colored neck.A bird taken in Februaiy is just as pale on neck and mantle as breed-ing specimens obtained in June. This race has a very limited dis-tribution, tlie most limited, with possibly one exception, of any formof the group.Seven specimens have been available, from the following localities,an asterisk indicating a breeding record
:
Lower California.?Espiritu Santo Island;* San Jose del Cabo.Measurements of specimens of Ardea herodias sanctilucae.
550 PROCEEDING.^ OF THE NATIONAL MUSEUM. VOL. 43.Type-locality.?Indefatigable Island, Galapagos Islands.Geographical distrihution.?Galapagos Islands: Indefatigable, Dun-can, Albemarle, and probably other islands of the archipelago.Doubtless permanently resident.Notwithstanding its far isolated range, this subspecies much moreclosely resembles the pale races of North America and the WestIndies than it does the very dark form of Mexico and Central Amer-ica. It differs from Ardea herodias treganzai in paler neck, stouterbill, and shorter tarsus; from Ardea herodias- wardi in lighter neck,stouter bill, and shorter wing and tarsus.One of the specimens examined (No. 116138, U.S.N.M., DuncanIsland) is not fully adult and is more brownish, somewhat darkerabove than the two adults. This race is doubtless peculiar to theGalapagos Archipelago, and probably occurs on many of the islands.Only three examples have been seen, from the following localities,breeding records being distinguished by an asterisk:Galapagos Islands.?Indefatigable Island; Duncan Island;* Albe-marle Island.*Measurements of specimens of Ardea herodias cognata.
Maseum and No.
U.S.N.M. 1890211
U.S.N.M. 11613S1
A.E.and O. Bangs,12451.
?
Sex.
[Male]
[Female]
...do..
Locality.
Al b em arleIsland ,GalapagosIslands. 2Duncan Is-land, Gal-apagos Is-lands.I n d efatiga-b 1 e I s-land, Gala-pagos Is-lands.2
Date.
Apr. 4, 1902
Apr. 13,1888
Feb. 16,1901
Collector.
R.H. Beck.
R. H. Beck.
NO. 1939. REVL^TON OF ARDEA HERODIAS?OnERHOLSEB. 55lface, including tail, tertials, innermost secondaries, and superior sec-ondary coverts, slate-gray, the long, narrow, plumaceous feathers of ?back and scapulars paler and glaucous, the median coverts and outerwebs of greater coverts paler gray, the rectrices more slaty on ter-minal portions; primaries, secondaries (except the innermost), pri-mary coverts, and alula, blackish slate, slate black or dull black;jugulum deep smoke gray, medially white, streaked broadly withblack and dull brownish slate, narrowly with cinnamon rufous andpale cinnamon, the long, narrow, pointed, plume-like feathers termi-nally pale drab gray, smoke gray, or whitish; a tuft of black, partlywhite-striped feathers on each side of the breast ; back of this a smallpatch of cinnamon rufous; sides and flanks slate gray; breast andabdomen black, broadly streaked with white; under tail-covertswhite; tliighs and edge of wing deep cinnamon njfous, the latter inpart more deeply colored, verging to chestnut; under wing-covertsslate color; axillars slate gray.Measurements.?Male:^ Wing, 488-511 (average, 497.0) mm.; tail,179-195 (186.5); exposed culmen, 139-148 (144.6); height of bill atbase, 27.6-32 (29.5); tarsus, 180-198 (188.3); middle toe, 109-119(112.4).Female:^ Wing, 460-492 (473.0); tail, 171-180.5 (175.5); exposedculmen, 135-140 (137.7); height of bill at base, 25-29 (26.5) ; tarsus,165-180 (170.3); middle toe, 94-109.6 (100.5).Type-locality.?liaird, northern California.Geogra'phical distrihution.?Pacific coast region of the United States,mainly in the Upper Austral and Transition zones: north to westernOregon; south to San Diego, southwestern California; east to SanGabriel, western California, and Baird, central northern California.Apparently a permanent resident throughout most if not all of itsI'ange, but wanders in winter west to the Farallon Islands, California,and east to St. John, Glenn County, California.This new race dift'ers much more from all the subspecies of Ardealierodias that are geographically near than it does from the far-distanttypical form of the species. It may readily be distinguished fromArdea lierodias treganzai by larger size and darker neck and upperparts, from Ardea lierodias sanctilucae by much darker neck andmantle and average longer wing. There does not seem to be anyconstant dift'erence in color between Ardea lierodias liyperonca andArdea lierodias lierodias, but the substantial difference in size serveswell enough to separate them.Two specimens from San Diego, California, taken respectivelyDecember 16, 1892 (No. 37088, Acad. Nat. Sci. Phila.), and April 24,1862 (No. 4494, Mus. Vert. Zool.^), the latter apparentl3''in the breeding
' Six specimens, from California.
- Three specimens, from California and Oregon.s Not seen by the writer, but eompared by Mr. Joseph Orinnell.
552 PROCEEDINGS OP THE NATIONAL MUSEUM. VOL. 43.season, are, in both size and color, essentially like other examples ofthe present formfrom northern California, although the first-mentionedbird is slightly paler, as would be natural from an area so near therange of Ardea lierodias treganzai.An immature winter example from St. John, Glenn County, Cali-fornia (No. 197862, U.S.N.M.), taken Januaiy 2, 1906, belongs appar-ently to this form; and the breeding bird of the whole Sacramento Val-ley may be the same. Two specimens collected by the United Statesexploring expedition under Captain Wilkes (Nos. 12670 and 15352,U.S.N.M.) and labeled simply "Oregon " must be referred to the pres-ent form, although rather darker above than birds from northernCalifornia. They are somewhat vergent toward Ardea lierodiasfannini,and probably came from near the mouth of the Columbia River?atleast no farther north?since all the great blue herons we have seenfrom Washington are Ardea Tierodias fannini.Specimens to the number of eight have been seen, from the localitiesgiven below:California.?Humboldt Bay; St, John; San Diego; Baird; SanGabriel.Oregon.?[No further locality specified.]Measurements of specimens of Ardea herodias hyperonca.
Museum and No. Sex. Lociility. Collector.
NO. 1939. REVISION OF ARDEA HER0DIA8?0BERH0L8ER. 553ARDEA HERODIAS OLIGISTA, new subspecies.CJiars. suhsp.?In color like Ardea herodias Jiyperonca, but lengthof wing very much less, bill more slender, and middle toe shorter.Description.?Type, almost adult male, No. 135573, U.S.N.M.;San Clemente Island, California, August 26, 1894; Dr. Edgar A.Mearns. Sides of crown, and occipital crest, black; center of crown,vertical crest, malar region, chin, and middle of upper throat, white;sides of head di-ab-gray ; neck all around deep smoke gray, the middleHne of foreneck streaked with black, v/hite, and cinnamon rufous;upper surface, including tail, tertials, innermost secondaries, andsuperior secondary coverts, slate gray, rather darker on middle ofback, the median coverts and outer webs of most of greater covertspaler gray, the rectrices terminally inclining to slate color or blackishslate; primaries, secondaries (except tlie innermost), primary coverts,and alula, blackish slate, slate black or dull black; jugulum smokegray, medially white, streaked broadly with deep smoke gray andslate color, anteriorly washed with cinnamon rufous, the long,narrow, pointed, plume-like feathers terminally pale gray or whitish;a tuft of black, partly white-striped feathers on each side of thebreast; back of this a patch of ciimamon rufous; sides and flanks slategray; breast and abdomen striped, black, slate gray, and white;lower tail-coverts white; thighs and edge of wing cinnamon rufous,the latter darker along secondaries, somewhat mixed with whiteat base of primaries; under wing-coverts slate color, with some edg-ings of chestnut; axillars slate gi"ay.Measurements.?Male:^ Total length (in flesh), 1160 mm.; wing,433; tail, 187; exposed culmen, 149; height of bill at base, 26; tarsus,184; middle toe, 101.Type-locality.?San Clemente Island, Santa Barbara Islands, Cali-fornia.Geographical distribution.?Upper Austral Zone on the Santa Bar-bara Islands, California: San Clemente Island; also Santa Cruz, SanNicholas, Anacapa, Santa Catalina, and probably other islands ofthe gi'oup. Apparently a permanent resident.Although the type is the only specimen, it is so very much smallerthan any example of the mainland form, Ardea Jierodias Jiyperonca,that it seemingly cam not belong to the same race. It has a shorterwing than even Ardea Jierodias herodias. From Ardea herodiastreganzai and Ardea Jierodias sanctilucae it differs in much shorterwing and darker neck and mantle.The type-specimen, though evidently full grown, is not quiteadult in color of plumage, which it shows in its brown forehead andfore part of cro\\m; in duller gray and in slight ochraceous admixture
1 One specimen, the type, from San Clemente Island, California.
554 PR0CEEDIN08 OP THE NATIONAL MUSEUM. vol. 43.on neck; some grayish in the tufts on each side of the breast, and onfeathers of breast and abdomen; and the not fully developed long,narrow, plumaceous feathers of back and scapulars.This race is probably confined to the Santa Bai-bara Islands, as thespecies is said to be resident there. Mr. G. WiJlett writes ^ that thegreat blue heron occurs and doubtless breeds in isolated pairs onmost, if not all of these islands, but that he has actually seen nests ononly Santa Catalina and Anacapa islands. Although, of course, weknow certainly of the subspecific status of only tlie bird on SanClemente Island, yet the published records of the great blue heronfrom Santa Catalina, Santa Cruz, and San Nicholas islands belongwithout much doubt under this form.ARDEA HERODIAS FANNINI Chapman.Ardea herodias fannini Chapman, -Bull. Amer. Mus. Xat. Hist., vol. 14, April 15,1901, p. 87 (Skidegate, Gi-aham Island, Queen Charlotte Islands, BritishColumbia).Chars, subsp.?Similar to Ardea lierodias liyperonca, but wing,exposed culmen, and middle toe shorter, the tarsus still more so; taillonger; upper parts and neck darker.Measurements.?Male:^ Wing, 472-492 (average, 480.7) mm.; tail,192-195 (193.7); exposed culmen, 124.5-137 (132.8); height of bih atbase,27-28 (27.3) ;tarsus, 153-167.5 (160.2) ;middletoe,97-107 (100.7).Female:^ Wmg, 456-486 (466.3); tail, 186-194 (189); exposed cul-men, 123-129.5 (126.8); height of bill at base, 24.5-26 (25.5); tarsus,146-158 (153); middle toe, 88.5-96 (93.2).Type-locality.?Skidegate, Graham Island, Queen Charlotte Islands^British Columbia.Geographical distribution
.
?Pacific coast region of northern NorthAmerica, mostly in the Transition and Canadian zones: north toHope, Cook Inlet, Alaska; south to Cape Flattery, northwesternWashington; and the Nisqually Flats, central western Washington.Permanently resident, except perhaps in the northernmost part of itsrange.This dark northwestern race is readily separable from Ardea herodiasherodias by its longer wing and tail, decidedly shorter tarsus, averageshorter culmen, much darker neck and upper parts. It is so verydifferent from Ardea herodias treganzai, by reason of its longer wingand tail, shorter tarsus, middle toe, and culmen, and very much darkerneck and upper parts, that close comparison is unnecessary. TheYBTj short tarsus of this subspecies, actually as well as relatively,distinguishes it from all the other forms of the species.
1 In letter.2 Three specimens, from Washington and British (Columbia.3 Three specimens, from Alaska and British Columbia.
NO. lO'JO. REVISION OF ARDEA HER0DIA8?0BERH0L8ER. 555An immature specimen from Admiralty Island, Alaska (No. 187748,U.S.N.M.), is very dark on neck and back, but this difference isapparently not more than individual. This race is confined, so faras known, to a narrow strip of country along; the coast and to theislands, from the State of Washington to Alaska.Ten specimens have been examined, from the sulijoined localities,breeding records being followed by an asterisk:Alaska.?Admiralty Island.British Columbia.^Snmas (Feb. 1); Sooke Lake; Skitlegate, Gra-ham Island (Queen Charlotte Islands) ; Victoria.Washington.?Cape Flattery; Nisqually Flats.*Measurements of specimens of Ardea herodiasfannini.
556 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.43.Tyi^e-locality
.
?Valley of Mexico, State of Mexico, Mexico.^
, GeograpMcal distribution.?Mexico, Central America, and northernSouth America, breeding in the Upper Austral, Lower Austral, UpperTropical, and Lower Tropical zones: north to southern Sinaloa;Colonia Garcia, northwestern Chihuahua; and Tamauiipas; south-east through all of Mexico to Guatemala, Nicaragua, Costa Rica,Panama, and Merida, Venezuela. Resident all the j^ear practicallythroughout its range, except perhaps in South America.From Ardea herodias Jiyperonca the present subspecies may bedistinguished by its shorter wing, darker neck and upper parts ; fromArdea lierodias herodias by longer wing and tail and darker neck andupper surface. The young in first plumage are like those of Ardeaherodias fannini, but are somewhat lighter on neck and upper parts.A single adult from Chihuahua City, Chihuahua, Mexico (No.187287, U.S.N.M.), taken February 22, 1904, incHnes very littletoward Ardea herodias treganzai; and a juvenal from Tamauiipas,exact locality unknown (No. 30477, Acad. Nat. Sci. Phila.), appearsalso to be A. h. lessonii. Specimens examined from Nicaragua andCosta Rica are all in juvenal plumage, and without adults it is diffi-cult to place them with certainty. They are no darker than juvenalArdea herodias herodias, and ixdght suggest the existence of anotherrace in this region, were it not that a juvenal of typical Ardea herodiaslessonii, from Patzcuaro, Michoacan, Mexico, is precisely the same incolor. It may be, therefore, that the young of Ardea herodias lessoniiis not darker than the same age of Ardea herodias herodias, althoughthe adult is decidedly so. At any rate, until more satisfactory ma-terial is available, the bird of Central America do^^'^l to Costa Rica,inclusive, must pass as Ardea herodias lessonii.A single bird in juvenal plumage from Empire, in the Canal Zone,Panama, is practically identical -with young birds from Alexico, Nic-aragua, and Costa Rica, except for being rather darker, more slaty,on the neck. Another bird, from Panama is, however, darker abovethan any adult specimen of Ardea herodias lessonii that we haveseen; and with, more material, the Panama bird may be separablesubspecificall}^ from that of Mexico. Since the above-mentionedexample is in rather an interesting phase of plumage, almost adult,and very much more brownish than either 3'oung or adult normallyare, the following description may be of interest:Almost adult male, No. 230144, U.S.N.M., Biological Survey col-lection; Fort Lorenzo, Panama, June 21, 1911; E. A. Goldman.Forehead and sides of sinciput cloA'e brown; sides of crown, andoccipital crest, black; superciliary stripe pale gray; center of crown,vertical crest, postocular region, malar region, chin and middle ofupper throat, white; subauricular region and sides of throat light
? Thus here restricted for the sake of definiteness.
NO. 1939. REVISION OF ARDEA HER0DIAS?0BERH0L8ER. 557brownish gray; neck all around deep brownish gray, somewhatpaler anteriorly, the median line of foreneck streaked with black,clove brown, cinnamon rufous, and white; upper surface, includmgtail, tertials, innermost secondaries, and superior secondary coverts,slaty brown, the middle of back darker, the median coverts and outerwebs of greater coverts lighter; primaries, secondaries (except theiimermost), primary coverts, and alula, brownish slate black; jugulumdeep brownish gray, medially white, broadly streaked \nth blackishslate, slate gray, and smoke gray, narrowly with chestnut and cinna-mon, the long, narrow, pointed, plume-like feathers terminally white;a tuft of brownish black, mostly white-striped, chestnut-streakedfeathers on each side of the breast; back of this a small patch of deepcinnamon rufous; sides and flanks brownish slate; breast and abdo-men white, broadly streaked with slate black and slate gray, nar-rowly with chestnut and cinnamon rufous; lower tail-coverts white;thighs and edge of wing cinnamon rufous, the latter partly mixedwith white ; lining of wing, including axillars, slate color, but some ofthe coverts with chestnut edgings.It thus differs from the fully adult bird in its brownish foreheadand sides of sinciput; deep brownish neck and upper parts; rustyedgings of upper wing-coverts; much mixture of white in the edgeof the wing; slaty tufts on the sides of the breast; mostly graybreast and abdomen; and the imperfect development of the dorsal,scapular, and jugular plumaceous feathers.From the date of capture of the previously mentioned Chihuahuaexample, and from other winter dates given in the Hst of specimensbelow, it is evident that the present race is resident throughout theyear over most if not all of its range.The subspecific name which is here applied has been entirely lostsight of for many years, and the writer is indebted to Mr. Ridgwayfor calling his attention to this long-forgotten description. ThisArdea lessonii of Wagler * is based on a great blue heron from
"Mexico," and there is no reasonable doubt concerning its properapplication to the present form.Twelve examples of this race have been seen, from localities asfollows, breeding records being indicated by an asterisk:Chihuahua.?Colonia Garcia (Feb. 22).Jalisco.?Ocotlan (Dec. 28); La Barca (Dec. 18).Mexico (State).?San Mateo (Dec. 9).Michoacan.?Patzcuaro .*Tamaulipas.?[No locality specified.]Costa Rica.?Lipurio.Nicaragua.?San Juan del Sur (Jan. 5).Panama.?Fort Lorenzo ;* Empire.
?Isis, 1831, p. 531.
658 PROCEEDINGS OF THE NATIONAL MUSEUM. VOL. 43.Measurements of specimens of Ardea herodias lessonii.
NO. 1939. REVISIO\ OF ARDEA HERODIAt^?OBERHOLSER. 559