cEP 27 1943 SARGENTIA A CONTINUATION OF THE CONTRIBUTIONS FROM THE ARNOLD ARBORETUM OF HARVARD UNIVERSITY IV MATERIALS FOR A FLORA OF THE CONTINENTAL NORTHWEST TERRITORIES OF CANADA BY A. E. PORSILD THE WILLOWS OF THE HUDSON BAY REGION AND THE LABRADOR PENINSULA BY HUGH M. RAUP WITH FOUR PLATES ee es ri ‘1 VER un . ii ~ R MT HAT ' ‘a ? il i TASH H } Y ; ay | TASH Ve . my ANE Nin mM CANN Raw v 3 SN _ 4 a PUBLISHED BY THE ARNOLD ARBORETUM OF HARVARD UNIVERSITY JAMAICA PLAIN, MASS., U. S. A. 1943 SARGENTIA A CONTINUATION OF THE CONTRIBUTIONS FROM THE ARNOLD ARBORETUM OF HARVARD UNIVERSITY A publication issued at irregular intervals by the Arnold Arbore- tum of Harvard University. Issues can be obtained from the Arnold Arboretum, Jamaica Plain, Mass., U. S. A. All correspondence pertaining to Sargentia should be addressed to the Librarian. No. I. Fiyran Prant Srupies, I]. Botanica, RESULTS OF THE 1940-41 CruisE oF THE “CHENG Ho.” By A. C. Smith (and collabo- rators). Pp. 1-148, with five text-figures. July 20, 1942. $2.50. No. I]. Tue ARracraceaz or Cuina. By Hui-Lin Li. Pp. 1- 134, with fourteen text-figures. Oct. 26, 1942. $2.25. No. III. A Revision oF THE GENUS SABIA COLEBROOKE. By Luetta Chen. With nine text-figures. Tur CHINESE AND. INDO- CHINESE SPECIES OF Ormosia. By E. D. Merrill and Luetta Chen. Pp. 1-120. Jan. 30, 1943. $2.00. No. IV. Martertats For A FLora oF THE CONTINENTAL NorTH- WEST TERRITORIES OF CANADA. By A. E. Porsild. Tur Writtows or THE Hupson Bay RecIon AND THE LaBRADOR PENINSULA. By Hugh M. Raup. With four plates. Pp. 1-135. Sept. 25, 1943. $2.50. SARGENTIA A CONTINUATION OF THE CONTRIBUTIONS FROM THE ARNOLD ARBORETUM OF HARVARD UNIVERSITY IV MATERIALS FOR A FLORA OF THE CONTINENTAL NORTHWEST TERRITORIES OF CANADA BY A. E. PORSILD THE WILLOWS OF THE HUDSON BAY REGION AND THE LABRADOR PENINSULA BY HUGH M. RAUP WITH FOUR PLATES a t iy y eo) s PUBLISHED BY THE ARNOLD ARBORETUM OF HARVARD UNIVERSITY JAMAICA PLAIN, MASS., U. S. A. 1943 SARGENTIA A CONTINUATION OF THE CONTRIBUTIONS FROM THE ARNOLD ARBORETUM oF Harvarp UNIVERSITY No. IV, pp. 1-135, with four plates Issued Sept. 25th, 1943 PRINTED BY THE LANCASTER Press, INc. LANCASTER, Pa. MATERIALS FOR A FLORA OF THE CONTINENTAL NORTHWEST TERRITORIES OF CANADA * A. E. Porsttp INTRODUCTION SoME of the earliest and most important botanical collections ever made in Canada came from the country between Hudson Bay and the Mackenzie River. This region was then known as Rupert’s Land, part of which is now included in the present Northwest Territories. By far the most important and extensive of these collections were made by the surgeon-naturalist and explorer Sir John Richardson, who, between the years 1819 and 1827, accompanied, as physician and naturalist, Sir John Franklin’s first and second expeditions “‘to the shores of the Polar Sea,” undertaken in the years 1819-22 and 1825-27. Later, 1848-49, Richardson led an expedition of his own in search of the missing Franklin Expedition of 1845. Large numbers of common Canadian plants and animals, until then unknown to science and to the rest of the world, were described from specimens brought back by Richardson and other members of the Franklin expeditions. It is but a fitting recognition of the work done that so many Canadian plants and animals are named in honor of such men as Richardson, Franklin, Drummond, Sabine, Hood, and Parry, to mention but a few of those intrepid pioneers in the scientific exploration of Canada, who, often under the greatest difficulties, personal danger, and hardship, brought back to the outside world specimens of the then unknown Canadian flora and fauna. A number of Richardson's plants were described and named by himself, others by Robert Brown in the botanical appendices to the reports of the Franklin expe- ditions. In all, Richardson reported 474 species of flowering plants and ferns, in addition to lower cryptogams, from what is now known as the Northwest Territories. In 1819 another British expedition, under W. E. Parry, in the Hecla and the Griper, wintered on Melville Island. The botanical collections made by Edward Sabine and other members of the expedition were described by Robert Brown in his famous “Chloris Melvilliana,” thereby laying the foundation for our present knowledge of the flora of Canada’s arctic islands. Between 1829 and 1840 ap- peared W. J. Hooker’s classic “Flora Boreali-Americana,” which until this day remains one of the standard handbooks on the flora of northern North America. As far as the continental parts of the Northwest Territories were concerned, information on the flora was based almost entirely on the results brought back by the Franklin Expeditions. In 1859 a young United States naturalist, Robert Kennicott, descended the Mackenzie River to the Peel, and crossed into Alaska. Owing to Kennicott’s premature death in 1866, a description of his small but important botanical collec- tion was never published. | * Publication of this paper was made possible by special funds provided for the purpose.— E. D. Merrill. 2 SARGENTIA : {4 In the years between 1887 and 1900 important contributions were made by members of the Geological Survey and the Topographical Survey of Canada: from the unexplored country between the Mackenzie drainage basin and Hudson Bay by J. B. and J. W. Tyrrell, and from the western shore of Hudson Bay by Robert Bell. A comprehensive account of the woody plants of the Mackenzie Basin was given by E. A. Preble in an appendix to his “Biological Investigation of the Athabasca-Mackenzie Region.” John Macoun, in his “Catalogue of Canadian Plants,’ completed in 1890, included the collections of the latter part of the nineteenth century, but for the geographical distribution of arctic plants depended in the main upon Hooker's Flora. Later additions to the flora of the Northwest Territories, including the important collections made on the west shore of Hudson Bay by J. M. Macoun in 1910, were published in appendices to reports of the Geological Survey of Canada. In 1906 appeared H. G. Simmons’ “The Vascular Plants in the Flora of Elles- mereland” and in 1913 his “Survey of the Phytogeography of the Arctic Ameri- can Archipelago.”’ These were based largely upon his work as a botanist with Sverdrup’s expedition in the Fram, 1898-1902, as well as upon the published and unpublished records of previous collections in the archipelago. In 1909 C. H. Ostenfeld published the important botanical results of Amund- sen’s Northwest Passage in the Gjoa, 1904-06. Since the appearance of these works, which had all dealt largely with the flora of the Arctic Archipelago, im- portant collections were made during the Canadian Arctic Expedition of 1913-18, chiefly by Frits Johansen and other members of the southern party under R. M. Anderson. This expedition collected plants along the mainland arctic coast from Alaska eastward to Bathurst Inlet and from the south shore of Victoria Island. The botanical results, by J. M. Macoun, Theodore Holm, Frits Johansen and others, were published in Volumes IV and V of the scientific report. Members of the Fifth Thule Expedition, 1921-24, under the leadership of Knud Rasmussen, collected plants chiefly in the Keewatin District. An account of their work has been pubished by Johs. Gréntved in the Report of the Fifth Thule Expedition, Vol. 2, No. 1, 1936. In 1940 appeared the first part of Nicholas Polunin’s ‘Botany of the Ca- nadian Eastern Arctic.” This important work brings up to date our knowledge of the flora of the Eastern Arctic, but the area covered by it extends only to the western shore of Hudson Bay. By far the most important contribution to our knowledge of the flora of the continental Northwest Territories since Hooker’s Flora is Hugh M. Raup’s “Phytogeographic Studies in the Athabaska-Great Slave Lake Region,” 1936, and “Botanical Investigations in. Wood Buffalo Park,” 1935, based largely upon his own very extensive collections made in the course of several expeditions to that region. In 1939 Raup spent a summer in the Mackenzie Mountains and his report on the flora of this very inaccessible and little-explored region is now in preparation. In 1927-28 the writer, with his brother, R. T. Porsild, while engaged in a grazing reconnaissance, collected plants in the northwestern part of the Mac- kenzie District, between the Mackenzie Delta and Anderson River, and in the Great Bear Lake basin.' The great distance to be covered, coupled with trans- 1A. E. Porsild. Reindeer Grazing in Northwest Canada. Department of Interior Publ. 1929, pp. 1-46. 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 3 portation difficulties and the very short arctic summers, made it impossible to do as much collecting as might have been desirable. Thus in one twelve-month period, in 1927-28, the writer moved his camp no less than 106 times. The collections of 1927-28 were added to materially during the years 1932-35, when the writer was stationed in the Mackenzie Delta. In 1931 the writer made a botanical survey of central Keewatin, from the Yathkyed Lake basin down Kazan River to Baker Lake. The vascular plants collected in the Northwest Territories during this and his earlier and later expe- ditions aggregate 5600 numbers, or over 30,000 herbarium specimens. In 1941 appeared the first part of Eric Hultén’s monumental ‘‘Flora of Alaska and Yukon.” The second part, though printed in 1942, has not been widely distributed due to the war. When completed, this will bring together all contem- porary knowledge of the flora of that region, which, because of the large land areas known to have been unglaciated there, must be considered the most impor- tant, floristically and phytogeographically, in the entire American Arctic. Only a few trained botanical collectors have ever visited the Northwest Terri- tories, and most of the collections of plants that have found their way into botani- cal museums and private herbaria have been made by travellers who have had no special botanical training and to whom the collection of plants was not the principal objective. For this reason nearly all species of aquatic plants, as well as some of the difficult and critical families and genera, such as Gramineae, Cy- peraceae, Salicaceae, Potentilla, Antennaria, Taraxacum, and others, are as a rule but poorly represented in their collections. Until the advent of the airplane all travellers in the north, at least during the summer, were compelled to follow the rivers and navigable canoe routes. Our present knowledge of the flora, therefore, on account of the past inaccessibility of vast areas, is still very incom- plete. Large tracts of land are still botanically unexplored, and the continental Northwest Territories even today remain among the least known large land areas in the northern hemisphere. Thus, while all available data on the floras of adjacent Alaska-Yukon, the Arctic Archipelago, the Eastern Arctic, and the Upper Mackenzie Basin have recently been brought together and published in the works of Hultén, Simmons, Polunin, and Raup, for most of the Mackenzie and Keewatin Districts, Hooker’s “Flora Boreali-Americana” still remains the principal published source of information. Since 1936, the writer has been engaged in the compilation and preparation of materials for a flora of the Mackenzie and Keewatin Districts. The principal modern source of material for this work has been in the collections made by himself and R. T. Porsild, supplemented by a number of small collections gath- ered by others and located in the National Herbarium of Canada. To this have been added a few data found in other herbaria, chiefly the Gray Herbarium of Harvard University, the herbarium of the New York Botanical Garden, and the herbarium of the University of Copenhagen, Denmark. This work was almost completed when the war stopped its final preparation and publication. In order to make some of the more important data available to other workers now engaged in studies of adjacent floras, it has been thought desirable at this time to publish, in a preliminary paper, descriptions of new species as well as the more important phytogeographical data. It is hoped that the main work, which will contain keys to the genera and complete data on dis- tribution, as well as copious ecological notes, may be published before long. 2A.E. Porsild. Geographical Journal 881; 1-19. 1936. a SARGENTIA [4 At present 806 species, subspecies, and major varieties of well-defined geo- graphical range are recognized in the flora of the Northwest Territories, while the number known to occur in the Mackenzie and Keewatin Districts is about 761. This number is made up as follows: Pteridophyta 27; Gymnospermae 6; Monocotyledoneae 213; Archichlamydeae 310; Metachlamydeae 205. Phytogeographically the delimitations of the Mackenzie and the Keewatin Districts are unnatural ones. The great Mackenzie valley is a strongly marked phytogeographic boundary, and the flora of the mountainous part of the Mac- kenzie District which lies to the westward of the valley bears a much closer af- finity to the flora of the Yukon Territory and Alaska than to that of the com- paratively low and dry peneplain east of the Mackenzie valley. Large parts of this plain are covered with spruce woods and muskegs and must be a formidable barrier to the migration of numerous species of plants. While, for practical reasons, the political boundaries are adopted to delimit most of the area covered by the present paper, a somewhat more satisfactory arrangement is achieved when the arctic sea-coast and coastal plain of the Yukon Territory are included with the Mackenzie District. This has been done in the present treatment. To the flora of the Mackenzie and Keewatin Districts have thus been added a few species that in Canada are restricted to the Yukon coast. Of the 761 plants constituting the present known flora of the Mackenzie (in- cluding the Yukon coast) and Keewatin Districts, 345 are enumerated in the present catalogue. Nearly one hundred of these are plants that have not previ- ously been recorded from, or that have not been known to occur in, the North- west Territories, while about 30 are unrecorded from the Mackenzie District only. In addition the following 14 are believed to be “new” to the flora of Canada: Carex rotundata Salix phlebophylla Carex rufina Melandrium taimyrense Tofieldia nutans Draba Palanderiana Salix arbutifolia Astragalus Collieri Salix Chamissonis Eritrichium aretioides Salix glacialis Senecio Kjellmanti Salix lingulata Taraxacum alaskanum Finally, in the writer’s collections were 21 species and 4 varieties believed to be new and undescribed. Several of these, in the following list marked with an asterisk, have been described elsewhere, while the rest are described here. *Potamogeton Porsildiorum Fern. Thlaspi arcticum n. sp. Calamagrostits chordorrhiza n. sp. Potentilla pulchella var. gracilicaulis n. var. Calamagrostis lapponica var. nearctica n. Oxytropis hyperborea n. sp. var. Gentiana Raupii n. sp. Poa ammophila n. sp. Antennaria crymophila n. sp. Kobresia arctica n. sp. Antennaria neoalaskana n. sp. Carex atrofusca var. decolorata n. var. *Antennaria philonipha A. E. Porsild Carex elynaeformis n. sp. Petasites arcticus n. sp. Carex Morrisseyi n. sp. *Taraxacum Carthamopsis M. P. Porsild Carex rariflora var. androgyna n, var. *Taraxacum spp.8 Melandrium Ostenfeldii n. sp. 8 Five or six species of Taraxacum in the writer’s collection have been named by Dr. Hag- lund, Lund, Sweden, but, as far as the writer is able to ascertain, the descriptions have not yet been published. 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 5 New combinations or names are proposed as follows: Melandrium Drummondii n. comb. Astragalus Colliert n. comb. Antennaria Ekmaniana n. nom. The remainder of the plants enumerated in the catalogue are critical entities in the flora or of unusual or little understood distribution. ACKNOWLEDGMENTS The writer is deeply indebted to a number of people who in various ways have assisted in this work. First of all he wishes to mention the large share of the collections contributed by his brother, Mr. R. T. Porsild of Dawson, Y. T., who in 1927 and 1928 accompanied him in the field. To his father, Dr. Morten P. Porsild, Director of the Danish Arctic Station, Disko, Greenland, the writer is deeply indebted for his stimulating and untiring interest in the progress and advancement of this work and for innumerable sug- gestions and notes on critical species, but above all for his painstaking compila- tion and preparation, in 1926, of a manuscript pocket flora containing brief de- scriptions and keys, and in many cases also pen drawings, of all species which, previous to 1926, had been recorded from arctic and boreal America. This pocket flora, in two volumes, proved of inestimable value in the field when, due to the great difficulties of transportation, none but the most indispensable hand- books could be carried. To the curators and staffs of a number of herbaria—the Gray Herbarium of Harvard University, Cambridge, Mass., the herbarium of the Arnold Arboretum of Harvard University, Jamaica Plain, Mass., the United States National Her- barium, Washington, D. C., the New York Botanical Garden, Bronx Park, New York, the Colorado Experiment Station, Colorado State College, Port Collins, Colorado, and the Botanical Museum of the University of Copenhagen, Den- mark—the writer is deeply indebted for making their material of arctic American plants available for study. The writer particularly wishes to express his gratitude to Professor M. L. Fernald and to Mr. C. A. Weatherby of the Gray Herbarium for valuable help in the solution of critical problems in taxonomy and nomenclature, as well as for ‘much stimulating encouragement during several prolonged and most profitable visits to that institution. Last, but not least, the writer is indebted to his friend Dr. Hugh M. Raup, of the Arnold Arboretum of Harvard University, whose own work on Arctic and boreal floras, like that of Professor Fernald, has been a constant source of inspi- ration, and who, over a period of years, has contributed numerous valuable suggestions. The writer received much assistance in the preparation of the Latin descrip- tions from Mr. A. W. A. Brown. The descriptions were later revised by Dr. Leon Croizat, of the Arnold Arboretum of Harvard University. THE CATALOGUE In the catalogue, serial numbers between 1861 and 5358, cited without col- lector’s name, are those of the writer and his brother; numbers between 5368 and 7473 are those of the writer alone; in all other cases the collector’s name has been given. Due to the necessity for conserving space, specimens are cited by serial number and the briefest possible data needed to identify the geographical position. 6 SARGENTIA INDEX TO MAJOR COLLECTIONS CITED Place and location Arctic CoAst oF YUKON TERRITORY Shingle Pt., lat. 69° N. ae Between King & Kay Pts. RicHarpson Mrs., west of Mackenzie Delta Slopes & foothills near Black Mt., between 1000 and 4000’ elev., approx. lat. 68° N. Lower slopes of Black Mt. Lower slopes between 1000 and 1800" elev. MackENZIE River DELTA Peel R., 20 mi. above Aklavik, Husky R. Peel R., Aklavik “ Limestone hills, Campbell Lake Campbell Lake East Branch, vicinity of Reindeer Station, wooded part about 68° 40’-50’ N. ae Lakes near Reindeer Station Reindeer Station Upper East Branch East Branch, near limit of trees Reindeer Station in a lake “é Lower East Branch Reindeer Station Lower East Branch Mouth of West Branch Lake between East Branch & Eskimo Lake South end of Richards Isl., 69° N. North end of Richards Isl., 69° 30’ N., 134° W. Hendrickson Isl. Kendall or Garry Isl. Kittigazuit, 69° 22’ N., 133° 40° W. Fish camp east of Kittigazuit Date July 12, 1933 Sept. 13, 1933 July 24, 1934 July 23, 1934 July 7-10, 1933 Aug. 15-17, 1933 Aug. 24-25, 1934 June 27, 1935 June 6-8, 1927 July 6, 1927 Sept. 1, 1934 June 15, 16, 1927 June 17-21, 1927 June 26, 1932 July 31, 1932 Aug. 7, 1932 Aug. 14, 1932 Sept. 4, 1932 July 30, 1933 Aug. 6, 1933 July 30, 1933 Oct. 16, 1933 June 20, 1934 July 12, 1934 July 21, 1934 Aug. 19, 1934 Aug. 10, 1934 Aug. 4, 1934 Aug. 18, 1934 Aug. 18, 1934 Sept. 10, 1934 Oct. 2, 1934 May-July, 1935 July 3-5, 1935 Sept. 16, 1933 Nov. 1, 2, 1933 July 22, 1934 July 18, 1927 July 22-24, 1927 July 26, 1935 July 26, 1935 December, 1927 July 19, 20, 1927 July 24-28, 1927 Arctic Coast, between Mackenzie and Anderson Rivers Tuktuayaktoq (Pt. Hearne) Atkinson Pt., 70° N., 131° 20’ W. Cape Dalhousie, 70° 20’ N., 129° 55’ W. Liverpool Bay, 70° N., 129° W. July 25, 1935 Aug. 1-3, 1927 Aug. 7-14, 1927 Aug. 15, 16, 1927 Serial Numbers 67 29-6730 6899-6911 7081-7115 7116-7208 6577-67 28 6742-6878 7316-7350 7398-7400 1861-1872 3181-3184 7302-7315 1873-1976 1978-2028 6500-6518 6519-6520 6521-0535 6536-6557 6558-6560 6731-6741 6879-6886 6887-6898 6912-6922 6940-6984 6985-7007 7008-7061 7255-7279 7211-7228 7229-7249 7250-7254 7280-7 299 7300-7301 7364-7371 7372-7380 7381-7397 6923-6924 6930-6939 7062-7080 2029-2130 2131-2296 7448-7455 7456-7463 3185-3187 2297-2411 2412-2530 7402-7447 2531-2681 2682-2812 2813-2951 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 7 Serial Place and location Date Numbers Eskimo LAKE BASIN Kugaruk, entrance to Lakes Aug. 31, 1927 3077-3095 2nd Lake, west end Aug. 19, 1927 2952-3018 Narrows between 3rd & 4th lakes Aug. 20, 1927 3019-3076 Setidgi Lake Aug. 21, 22, 1927 3096-3180 MACKENZIE RIVER Bear Rock, below Norman July 23, 1928 3375-3399 Ramparts above Good Hope Sept. 19, 1928 3402-3405 : Sept. 25, 1928 7472-7473 Mackenzie R. between 62° 20’ and 65° 40’ N. Sept. 17-27, 1931 6561-6573 GREAT BEAR LAKE Bear River June 15, 1928 3238-3274 Mt. Charles July 16, 1928 3275-3321 Sept. 2, 1928 3322-3370 Keith Arm, Ft. Franklin May 11-June 13, 1928 3188-3237 a Russel Bay Aug. 26-28, 1928 3407-3467 Etacho Point, 1500’ elev. Aug. 24, 1928 3468-3524 Smith Arm: north shore July 16-26, 1928 5054-5087 Dease Arm: north shore June 23-26, 1928 4661-4730 Barrens near headwaters of Horton River June 19-21, 1928 4731-4751 Foot of Dease Arm June 7—Aug. 1, 1928 4752-4882 North shore June 27-July 22, 1928 4883-5053 Cape McDonnel Aug. 1, 1928 5088-5164 McTavish Arm: north shore Aug. 46, 1928 5165-5252 Head of Edna Travers Bay Aug. 8, 1928 5253-5285 East coast Aug. 8-11, 1928 5286-5350 Southern portion Aug. 13-21, 1928 3599-3786 Leith Point Aug. 23, 1928 3525-3598 Miscellaneous small collection 1928 6182-6200 KEEWATIN DIstRICT Lake on Tha-anne R., 60° 58’ N., 97° W. July 12, 13, 1930 5544-5607 West Coast of Hudson Bay, Mistake Bay, 62° 05’ N. July 20-29, 1930 5608-57 30 Yathkyed Lake on Kazan R., 62° 30'-63° N., 97°- Aug. 1-15, 1930 5731-5861 98° 30’ W. Lower Kazan R. Aug. 16-26, 1930 5992-6074 Baker Lake, 64° 30’ N., 97° W. Sept. 2, 3, 1930 6075-6129 Chesterfield Sept., 1930 6130-6175 MISCELLANEOUS COLLECTIONS Chesterfield Inlet, Dr. Currie (Polar Year Exp.) 1-72 Yellow Knife Reserve, 63°-64° N., 109°-114° W., 1936, John Carroll 1-25 Barrens between Backs and Thelon R., 64° 20’ N., 104° W., John Carroll 26-67 Thelon Game Sanctuary, 1936, C. H. D. Clarke 1-116 Coppermine R., 1931, A. M. Berry 1-28 Foothills west of Mackenzie Delta, 1931, O. Bryant unnumbered The symbols mentioned below indicate the herbaria in which the specimens cited were seen. Where no symbols are appended, the specimens are in the National Herbarium of Canada. C—Herbarium of the University of Copenhagen, Denmark Can—Herbarium of the National Museum of Canada, Ottawa G—Gray Herbarium, Harvard University, Cambridge, Mass. NY—Herbarium of the New York Botanical Garden, New York, N. Y. US—United States National Herbarium, Washington, D. C. The families and genera are arranged essentially according to Dalla Torre and Harms, Genera Siphonogamarum (1900-1907). Within the genera the 8 SARGENTIA (4 species, for the sake of convenience, are arranged alphabetically, except in the genus Carex, where the arrangement followed is that of Mackenzie in N. Am. Flora 18 (1935). EQUISETACEAE Equisetum palustre L. MACKENZIE River Dertta: Aklavik, 3181; Great Bear Lake: North shore of Smith Arm, 4981; McTavish Arm, Conjuror Bay, 4858. Recorded by Richardson “‘to the shores of the Arctic Sea,” but not otherwise known from the Northwest Territories. SELAGINELLACEAE Selaginella selaginoides (L.) Link. GREAT Bear Lake: Keith Arm, Russel Bay, 3416, 3433; south shore, Leith Pt., 3563; Dease Arm, north shore, 4662. New to the flora of the Northwest Territories. Selaginella sibirica (Milde) Hieron. Ricuarpson Mrs.: East slope, west of Mackenzie Delta, about 68° N., 2000-4000’ elev., 6633, 6746; Mackenzie River Detta: East Branch, 68° 40’ N. to 68° 55’ N., 6559, 6574, 7008; south end of Richards Isl., 2031. New to the flora of the Northwest Territories and not previously recorded from north of Lake Athabaska. ISOETACEAE Isoétes Braunii Dur. Great Bear Lake: McTavish Arm, (plants with mature macrospores) 3688. In the Mackenzie Basin J. Braunii was previously collected on Lake Athabaska (Raup & Abbe 4617). It is thus new to the flora of the Northwest Territories. POTAMOGETONACEAE Potamogeton pectinatus L. MackeENzIeE River Detta: Lake near the East Branch, 68° 40’ N., 6913, 6915, 7244. The last number, collected on Aug. 11, 1934, has immature fruit, showing that in the Mackenzie Delta the species produces fruit at least in favorable seasons. Potamogeton pectinatus in Canada was known previously north to Lake Athabaska. SPARGANIACEAE Sparganium angustifolium Michx. Eskimo Lake Basin: Setidgi Lake, 3097, 3173 (the last with mature fruiting heads) ; Great Bear Lake: McTavish Arm, 5334. Sparganium simplex, in Hook. Fl. Bor.-Am. 2: 169. 1839, “to Ft. Franklin,” may belong here, but otherwise the species has not been previously recorded from north of Great Slave Lake. Sparganium hyperboreum Laest. Mackenzie River De_ta: West Branch (Husky River), 7316; Eskimo LAKE BasIN: 2954, 3059, 3125; Great Bear LAKE: 3434, 4983, 5056; Krewatin Distr.: Yathkyed L., 5732, 5998; Tha-anne R., 5550; Churchill, Mosquito Pt., 5521. Apparently common and in favorable seasons maturing fruit north to, or slightly beyond, the limit of trees. 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 9 Sparganium eurycarpum Engelm. Mature fruiting heads were collected by the writer on the Mackenzie River, near Ft. Norman, in September, 1928. In the Mackenzie basin not previously recorded from north of Lake Athabaska. SCHEUCHZERIACEAE Triglochin maritima L. Mackenzie River Detta: Campbell L., 1899; Kittigazuit, 2301; Great Bear Lake: 3261, 3671, 4664, 5089, 5165. Triglochin palustris L. Mackenzie River Detta: Recorded in many places (A. E. Porsild, field notes) ; Great Bear LAKE: 4664A, 4785, 5166; KEEWATIN Distr.: Mistake Bay, 5612. The above numbers represent the first collections from the Mackenzie basin north of Great Slave Lake, and the first from Keewatin District. GRAMINEAE Beckmannia Syzigachne (Steud.) Fern.; Raup in Jour. Arn, Arb. 17: 209, 1936. Beckmannia erucaeformis R. & S.; Richardson in Franklin’s Journ. App. 731. 1823. Although Richardson apparently collected this plant, Hooker, Fl. Bor.-Am., has not included it. Richardson does not give any stations but merely lists it under “W” (‘wooded country from latitude 54° to 64° north’), and Raup’s, 1. c., seems to be the first specific record from the Northwest Territories (Miss E. Taylor 30099 [Can], Ft. Simpson). Along the lower Mackenzie River, B. Sysigachne occurs sporadically near settlements as a weed introduced from the upper river with native hay shipped north with livestock. In the Mackenzie Delta, in addition, there occurs a plant which apparently is indigenous, but of a very curious local distribution. It inhabits muddy shores of certain shallow, stagnant lakes flooded during the early part of the season. But even in such places its occurrence is sporadic (Mackenzie River Delta: East Branch, 6898 and 7235). The seeds are eagerly eaten by wild ducks, and the fact that the plant grows in places frequented by these birds suggests that the latter distribute the seeds. Were the seeds carried by flood water the distribution would, no doubt, be more general. Schizachne purpurascens (Torr.) Swallen. Great Bear Lake: McTavish Arm, Conjuror Bay, 3643. New to the flora of the Northwest Territories. Muhlenbergia Richardsonis (Trin.) Rydb. Great Bear Lake: Bear River, 3357. New to the flora of the Northwest Territories. Calamagrostis chordorrhiza sp. nov. Gramen perenne laxe caespitosum, rhizomatibus longis sublignosis subnitidis perramificatis ; culmi ascendentes aut geniculati, 30-35 cm. alti, omnino scabri, solitarii aut aggregati, phyllopodici, binodosi; folia inferiora caulina numerosa, aggregata, brevia, laminis planis latisque instructa, superioribus angustioribus valde reductis; folia axium sterilium numerosa, in fasciculis foliosis conferta, culmo maturo dimidio breviora; lamina 3 mm, lata, plana, marginibus parce in- volutis abrupteque acuminatis, glauca, utrinque scaberrima; ligula triangularis, 2 mm. longa, erosa; panicula laxa angusta, 5-6 cm. longa et 1-1.5 cm. lata, + interrupta; rami breves hispidissimi, inferioribus remotiusculis; glumae aequales, 2.5-3 mm. longae, acuminatae vel fere cuspidatae, parum purpuras- 10 SARGENTIA |4 centes ; carina subhispida, marginibus vix puberulens, marginibus angustis pel- lucidis instructa; flosculus 2-2.3 mm. longus, lemmae firmiusculae infra virides, apicibus stramineis, arista recta vel rarissime subgeniculata, quam lemma quinto minor, evidenter infra medium orta, palea pellucida, lemmam aequans; pili ad basim paleae inaequales, copiosi, longissimi aequi ac lemma tantum longi; rha- chillae nodo brevissimo. Mackenzie River Detta: East Branch, 68° 40’ N., hills back of Government Reindeer Station, forming large cushions in sandy soil, Aug. 7, 1932, 4. E. Porsild 6524 (type). Thus far known only from the type locality. Calamagrostis chordorrhiza, because of its small spikelets and small, straight awn, may be related to C. neglecta (Ehrh.) Gaertn., while its very scabrous and glaucous leaves and strong rhizomes are somewhat suggestive of C. purpurascens. Our plants are late-flowering but the floral parts appear quite normal and do not suggest hybrid parentage. By its fascicles of flat, crowded and glaucous leaves and strong, cord-like rhizomes, C. chordorrhiza seems so well-marked from all other species of Calamagrostis known to the writer that he can see no other alternative than to consider it undescribed. Calamagrostis deschampsioides Trin. Although Polunin (in Nat. Mus. Can. Bull. 92: 50. 1940) was the first to record this species from Canada, the writer and his brother, in 1927, collected it in the Mackenzie Delta (north end of Richards Island, July 22-24, 1927, 2136, 2137; Arctic Coast: Atkinson Point, 2558). In the writer’s collection also are specimens from Churchill, Man., 5378. Calamagrostis lapponica (Wahlenb.) Hartm. var. nearctica var. nov. Calamagrostis lapponica (Wahlenb.) Hartm.; Britton & Rydberg in Bull. N. Y. Bot. Gard. 2: 154, 1901. Calamagrostis canadensis var. Langsdorffii sensu Porsild in Rhodora 41: 179. 1939, pro min. parte, non Link. A specie differt statura semper minore (35-50 cm. neque 50-120 cm.), pa- nicula breviore angustiore (4-8 cm. neque 5-15 cm.), glumis parum brevioribus angustioribusque (4 mm. neque 5 mm.). The writer, in 1927 and 1928, collected in various parts of the Mackenzie District what he then took to be a slender form of C. canadensis var. Langsdor ffi. From 1932 to 1935, when he was a resident of the Mackenzie Delta, this plant was grown from seed in its natural habitat and also studied in the field. It was realized then that the plant did not belong with the polymorphous C. canadensis, and he at first thought it to be an undescribed species widely distributed in arctic and subarctic parts of Northwest America, since clearly it could not be referred to any American species. Comparison with European and Asiatic species has shown, however, that it must be referred to C. lapponica (Wahlenb.) Hartm., although it differs so consistently from that species in size that it seems best treated as an American (? and E. Asiatic) race or variety. The following is a more detailed description. Loosely caespitose with short, thin rhizomes; culms solitary or few together, slender, erect, 30-60 cm. high, glabrous except just below the panicle, with 2 or 3 nodes, the cauline leaves strongly involute; leaves about half as long as the mature culms, about 2 mm. wide, flat, or more often strongly involute, essentially glabrous; ligule short and entire, almost truncate, 1.5-2 mm. long, sometimes tinged with purple; panicle lax, suffused with purple, narrow to somewhat open, 4-6-8 cm. long and 1.5, rarely 2, cm. wide, the branches very hispid, the 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 11 longest 2.5 cm. long; glumes 44.5 mm. long, slightly lustrous and. strongly suffused with purple, with stramineous margins and tips, abruptly acuminate, hispidulous-scabrous on the keel and sparsely scabrous on the sides; floret 3.6-3.7 mm. long, the lemma truncate, 7-nerved (counting mid-nerve, which is wanting above the base of the awn), scabrous except at the hyaline summit; awn arising 14 from the base, erect, straight or rarely slightly bent in the lower half, barely equaling the lemma; palea 2.7 mm. long, erose-truncate, the margins and upper half hyaline, the nerves scarcely reaching to the apex; callus hairs unequal, the longest nearly as long as the lemma; rhachilla joint 1 mm. long, its hairs copious, as long as the lemma, a few slightly longer. In old sand dunes or in peaty soil or in open heath and muskegs, on the arctic coast from Norton Sd., Alaska, east to Hudson Bay, south to Great Bear Lake and mountains of central Yukon and Alberta. MackenzeE River Detta: East Branch, 68° 40’ N., two-year-old plants grown from seed (collected with specimens of no. 6525), Aug. 12, 1934, A. E. Porsild 7282 (tyPE). Cala- magrostis lapponica var. nearctica, in addition, is represented by the following specimens in the National Herbarium of Canada: ALasKka: Norton Sd., Pastolik, 955 (distributed as C. canadensis var. Langsdorffii) ; YUKON Territory: Lake Kluane to Don Jek River, Aug. 11-27, 1920, Adolf Miiller (distr. as C. scabra Presl) ; Klondike, Indian Divide, John Macoun 54626 (distr. as C. neglecta Ehrh.) ; same place, John Macoun 54627 (distr. as C. canadensis Michx. var. acuminata Vasey) ; MACKENzIE District: Mackenzie River Delta, East Branch, 68° 40’ N., solitary culms in dry heath, Aug. 7, 1932, 6525 (topotyPE) ; Eskimo Lake Basin, north shore of 2nd lake, low, marshy tundra, 2959, 2960; Great Bear Lake, Keith Arm, Rus- sel Bay, open willow thickets, 3408; same place, sand dunes along lake shore, 3419; KEEWATIN District: Yathkyed L. on Kazan River, dry hillsides, 5738; Baker Lake, A. Dutilly 4513 (distr. as C. purpurascens) ; Chesterfield, west coast of Hudson Bay, dry tundra, 6132. The following, from high mountains of Alberta, differ chiefly in slightly smaller (3 mm.) and more glabrous glumes: AtBerta: Mountain Park, Mt. Harris, 7000 ft. elev., Aug. 13, 1925, Malte & Watson 2134, 2153; back of “Miner’s Roof,” 6000 ft. elev., Aug. 11, 1925, Malte & Watson 2017. Calamagrostis lapponica var. nearctica may be distinguished from all other N. American species of Calamagrostis by its narrow, lax panicle, which is strongly suffused with purple, its short and often involute leaves, and by its solitary culms, its long callus hairs, and straight awn. On the other hand it is a good match for C. lapponica, Fl. Dan. Suppl. 1, tab. 5 (1853), as well as for specimens in the National Herbarium of Canada thus labeled, exemplified by: “Pl. Finl. Ex- sicc. No. 456, Lapponia Kemensis, Aug. 2, 1913, leg. U. Segerman” and “Montes uralenses septentrionales (lat. bor. inter 64° et 65°), July 27, 1927, leg. V. Soc- zava, no. 386,” but it is consistently smaller than the plant called for in the original description as well as the plant described by modern European authors such as Holmberg (Skandinaviens Flora 11: 153. 1932) and Roshevitz (in Ko- marov, Fl. U. R. S. S. 2: 219. 1934). The former gives the following extra- Scandinavian distribution: ‘‘N. Russia; Siberia; N. America; Greenland.” Holmberg’s American record may be based upon a note by Fernald in Rhodora (13: 123. 1911), and the Greenland one undoubtedly refers to C. lapponica Hartm. var. groenlandica Lange (Conspectus Fl. Groenlandica 2: 296, 1887), which is C. hyperborea Lange. Calamagrostis lapponica has from time to time been recorded from North America (Britton & Rydberg, 1. c.; Rhodora, 1. c.; St. John in Vict. Memorial Mus. Mem. 126: 64. 1922; Stebbins in Rhodora 32: 56, tab. 195, fig. 9. 1930; Victorin, Fl. Laurentienne 792. 1935), but all but the record of Britton and Rydberg seem very doubtful, and all plants thus labeled, seen in America her- 12 SARGENTIA [4 baria by the writer, seem best referred to C. neglecta (Ehrh.) Gaertn. or to C. hyperborea Lange. Certainly C. lapponica var. brevipilis Stebbins (Rhodora, l. c.) should be referred to C. neglecta (Ehrh.) Gaertn. In his revision, Stebbins, 1. c., appears to have overlooked the plant noted as C. lapponica (Wahlenb.) Hartm. by Britton and Rydberg, |. c., from the Yukon. Their description and discussion clearly show that their plant is C. lapponica var. nearctica, and it thus appears that theirs is the first authentic record of C. lap- ponica from North America. Trisetum sibiricum Rupr. Trisetum flavescens Ostenf. in Gjéa Exp. 32. 1910. Ostenfeld, |. c., “with some doubt,” recorded this species from King Point on the Arctic coast of Yukon Territory, stating that the specimens were very poor and much too young. The writer succeeded in finding mature specimens, prob- ably in the identical place: Between King and Kay Points, 69° 12’ N., 138° 30’ W., July 23-25, 1934, 7122. . Poa ammophila sp. nov. Gramen perenne dense caespitosum, rhizomate brevi colonias densas compactas formante ; culmi 20 cm. raro 35 cm. alti, erecti vel parum ascendentes, rigidi, perglabri, folia circa duplo superantes, foliis 1 aut 2 in parte inferiore muniti: folia glabra, cinereo-viridia rigida, involuto-conduplicata; vaginae subinflatae, persistentes, nitidiusculae, ligula acuminata, circa 1.5-2 mm. longa; panicula etiam tempore anthesi angusta, circa 4 cm. longa et 0.5-0.8 cm. lata, ramis bre- vibus appressisque, glabris vel scabriusculis; spiculae circa 6 mm. longae, ple- rumque triflorae, haud conspicue compressae; glumae aequales, circa 4° mm. longae, strictae, carinatae, parte superiore tertia glabrae aut subscabrae, sub- nitidae, purpurascentes, margine late scariosae, acuminatae ; lemmae circa 3.5 mm. longae, forma glumas simulantes, dorsaliter brevi-pubescentes, basi haud ciliata obscure uninervatae, membranaceae ; antheris circa 1.4 mm. longis. Arctic Coast: Cape Dalhousie, 70° 20’ N., 125° 55’ W., forming colonies on sandy hills back of coast, Aug. 7-14, 1927, 2704 (type); 6 miles east of Kittigazuit, 2422: Cape Dal- housie, a luxuriant form 35 cm. tall, 2706; Liverpool Bay, 2827; Mackenzir River Devta: North end of Richards Isl., 2154; Kittigazuit Island, 2311. In no. 2827 the spikelets are less flattened, the glumes are shorter and less prominently keeled, and the ligule is somewhat shorter than in the type. Poa ammophila is a characteristic plant of old sand dunes and sandy hilltops along the arctic coast from the Mackenzie Delta east to the mouth of Anderson River. Poa ammophila superficially resembles P. Hartzii of Greenland, recently, with some doubt, reported also from Ellesmere Island (Polunin in Nat. Mus. Can. Bull. 92: 69. 1940), but the abundant material on hand all differs from it in the much shorter ligules and the much shorter and less abundant pubescence of the lemmas. In fact it seems very close to some Cordilleran species, and, following Rydberg, Fl. Rocky Mountains, it seems best placed in the section Rupicolac, where it most closely matches P. Pattersoni Vasey. This, however, is said to have a short, truncate ligule. Except that the glumes are longer than the lemmas, it matches the original description of that species and also a specimen from the type locality: Gray’s Peak, Colorado, 13,000 ft., Aug. 15, 1885, G. W. Letterman (Can). It is also a close match for Poa Suksdorfii Vasey, Mt. Rainier, Wash., Aug. 14, 1895, O. D. Allen 183 (Can). Our plant, however, is a good deal taller and also differs so much from half a dozen other sheets in the National Her- barium of Canada named P. Pattersoni by Rydberg and others that the writer hesitates to refer it to that species. 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 13 Festuca altaica Trin. RicHarpson Mrts.: West of Mackenzie Delta, 68° N., 3000-4000’ elev., 6636; GrEAT BEAR LaKkeE: Dease Arm, 4667; north shore of Smith Arm, 5070; Arctic Coast oF YUKON TERRI- tory: Near Shingle Pt., 6902, 7083. New to the flora of the Northwest Territories. Agropyron latiglume (Scribn. & Merr.) Rydb. Agropyron alaskanum Scribn. & Merr., Macoun & Holm in Rept. Can. Arct. Exp. 5A: 8. 1921. Occasional in sandy or somewhat peaty soil from the mountains west of the Mackenzie Delta, east along the Arctic Coast and on islands of the Arctic Archi- pelago; Kazan River, south of Baker Lake (no. 6000), and south to Great Bear Lake and the east end of Great Slave Lake. Also in Yukon Territory and Alaska. Agropyron trachycaulum (Link) Malte. MACKENZIE River Detta: East Branch, 68° 40’ N., 6519; south end of Richards Isl., 2048; 6 miles east of Kittigazuit, 2416; Great BeAr LAKE: Dease River valley, 4864; north shore of Smith Arm, 5067; north shore of McTavish Arm, 5195. Agropyron trachycaulum in its distribution is more southern than A. latiglume and in the Northwest Territories does not reach far beyond the limit of trees. It seems always restricted to alluvial soils. On clay banks of the East Branch of the Mackenzie Delta it forms large, firm tussocks, one meter or more in diameter and often 1.25 m. high, and has spikes 16 cm. long. Our material is all rather uniform. The lemmas are pubescent with a short awn and the sheaths are glabrous. According to Malte’s revision (Ann. Rep. Nat. Mus. Can. 1930: 48. 1932) this would place it with var. pilosiglume Malte, known only from the type locality, Vancouver Island, B. C., which somehow does not make sense. Elymus innovatus Beal. Elymus mollis R. Br., sensu Hook. Fl. Bor.-Am. 2: 255. 1840, not Trin., at least in part. RIcHARDSON Mrs.: Foothills west of Mackenzie Delta, 6634; MACKENZIE RIVER DELTA: E. Branch, 68° 40’ N., 6501, 6521; MackeNzie Riyer: Bear Rock, 1300’ elev., 3375; Great Bear LAKE: Bear River, Mt. Charles, 3324, A western species restricted to calcareous soils of the interior; in the Mac- kenzie country it barely reaches north to the limit of trees. CYPERACEAE Eriophorum callitrix Cham. Since Fernald (in Rhodora 27: 203-210. 1925) cleared up the true identity of E. callitrix and its distribution in North America, our knowledge of the known range of this plant has been enormously increased. Polunin’s prophetic state- ment (1. c., 102) that “when explorers have learned to overlook it less it may prove circumpolar’” may indeed some day come true. It has already turned up in East Greenland, and its discoverer, G. Seidenfaden, with Th. Sdrensen, has since published a paper on the geographical distribution of it and allied species (Medd. om Grgnland 1011: 1-27. tab. 1, 3, 1933). To the latest new stations in Alaska recently given by Porsild (in Rhodora 41: 199. 1939) and in the Eastern Arctic by Polunin, |. c., should be added the following : RicHarpson Mrs.: West of Mackenzie Delta, 2000-3000’ elev., 6639; Great Bear LAKE: Foot of Dease Arm, 4753; north shore, Haldane R., 5026; Keewatrn Distr.: Tha-anne R., 60° 58’ N., 97° W., 5570; James Bay: S. Twin Isl., 4205. Also, in the Gray Herbarium, is a specimen from Montana: Custer National Forest (L. O. Williams 3706). E. callitrix ap- pears to be restricted to calcareous soils. 14 SARGENTIA 4 Eriophorum medium Anders. Mackenzie River Detta: East Branch, 6958; S. end of Richards Isl., 2050; KrEEwatin Distr.: Tha-anne R., 5570B; Yathkyed Lake, 5755, 5756; Kazan R., 6018. Eriophorum opacum (Bjérnstr.) Fern. MaAcKENZIE River Detta: Peel R., above Aklavik, 7400; Campbell Lake, 1885. Eriophorum opacum in the north is the latest-flowering species in this genus and is confined to muskeg forest. Its range is decidedly more southern than that of the other caespitose species. Where Seidenfaden and S¢rensen, |. c., give the world range of E. opacum, the reader is astonished to find that among other stations under the heading “Arctic America” are “Hastings County, On- tario and Prince Albert, Sask.,” and that Churchill, Manitoba, has been placed in the Northwest Territories. In fact the reader might easily be misled to be- lieve, with Seidenfaden and Sgrensen, that the American range of FE. opacum is arctic. Eriophorum vaginatum L. The true £. vaginatum of Eurasia is a very common species in the northern parts of the Mackenzie District, where it is a dominant species in “climax” tundra. The species in the Mackenzie Delta flowers late in May. Late snow or rain during anthesis may completely ruin the season’s seed crop; but otherwise, during late July, large tracts of land, or sometimes the entire landscape, turn completely white as if covered with snow. Heads composed entirely of pis- tillate flowers are as common as the normal bisexual ones. The anthers are from 2.1 to 2.6 mm. long, the achenes flat-trigonous, almost black, 2-3 mm. long, and 2 mm. wide. East of the Mackenzie Delta E. vaginatum becomes progressively less ubiqui- tous, and at Great Bear Lake it no longer dominates. It is still less common in Keewatin District and east of Hudson Bay, and in the southern parts of the Mackenzie District it is replaced by the very closely related FE. spissum Fern. Eleocharis acicularis (L.) R. & S. MAckKENZIE River DELTA: Forming masses of vegetation in nearly all lakes (var. sub- mersa (Hj. Nilss.) Svenson), while the fertile plant grows abundantly on wet, muddy shores; East Branch, 6918, 7236, 7410, the last number fertile and grown in the writer’s garden on the East Branch from seed; Arctic Coast: East of Kittigazuit, 2452 (var. submersa) ; same place, 2432 (fertile) ; Great Bear Lake: Dease Arm, 4669, 4789; Smith Arm, 4987; Mc- Tavish Arm, 3691; KreEwatiIn Distr.: Yathkyed Lake on Kazan R., 5749 (fertile), 5750 (var. submersa) ; Lower Kazan R., 6019; Baker Lake, 6180 (sterile) ; Mistake Bay, west shore of Hudson Bay, 5623 (sterile). ‘ Not previously recorded from north of Great Slave Lake. Eleocharis palustris (L.) R. & S. Great Bear Lake: North shore of Dease Arm, 4668; Great Bear River, 3376. Previously recorded from Bear Lake by Richardson. Eleocharis pauciflora Link. Great Bear Lake: North shore, Haldane R., 4914; Dease Arm, 4670; Smith Arm, 4987 A; McTavish Arm, 5170; south shore, Leith Pt., 3529. Not previously recorded from north of Great Slave Lake. Scirpus validus Vahl. MaAcKENZIE River: Near Ft. Norman, 5367. Only record north of Great Slave Lake. Kobresia caricina Willd. Kobresia bipartita (All.) Dalla Torre. 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 15 Common or occasional in rather dry tundra and represented by many numbers in the writer’s collection, from the Mackenzie Delta east to Hudson Bay. Not previously collected north of Great Slave Lake. Kobresia arctica sp. nov. Planta densissime caespitosa, colonias parvas sed compactissimas formans raro 15 cm. diametro superantes; rhizomata brevissima, ascendentia, crassa, non- stolonifera ; culmi 12-35 cm. alti, flavo-virides, rigidi, obtusanguli, leves, maturi folia multum superantes, phyllopodici vaginis annorum praecendentium siccis persistentibus valde obviis; folia erecta, longe attenuata, opaca viridia, angusta, valde canaliculata, facie teretia, marginibus parte tertia inferiore scabriusculis ; vaginae opacae castaneae ; spica solitaria, ovata vel ellipsoidea, ebracteata, 1-1.8 cm. longa ac 0.5-0.8 cm. lata; spiculae circa 20, unaquaque florem masculum simul cum foemineo continente duobus infimis exceptis in rhachillis vestigialibus impositis, flores 2 vel 3 foemineos, masculum 1 continentibus; squamae_per- sistentes, late ovatae, obtusae, opacae, obscure castaneae centro pallidiores, mar- ginibus angustis pellucidae; nuculae 3-4 mm. longae, utriculis chartaceis per- sistentibus breviores, obovoideae, obvie triangulares, pallidae brunneo-cinereae, glabrae, opacae, enervosae, brevistipitatae fere erostres; stylus tenuis nuculam subaequans ; stigmatibus 3. In rather dry, peaty tundra from mountains west of the Mackenzie River east to Back’s River and south to Great Bear Lake. Also islands of Bering Sea and in eastern Siberia (?). MACKENZIE River Detta: Kittigazuit Island, 69° 22’ N., 133° 40’ W., common in dry tundra, July 19, 20, 1927, A. E. & R. T. Porsild 2318 (type); Rricnarpson Mrs.: West of Mackenzie River Delta, in peaty soil, alpine tundra, 3000’ elev., approx. 68° N., 136° W., 6757; Arctic Coast oF YUKON Territory: 7125; MACKENzIE River DELTA: Mainland near reindeer corrals, 7381; Hendrickson Isl., 7456; Kittigazuit Isl., 7371; north end of Richards Isl., 2159; Arctic Coast: Atkinson Pt., 2562; Cape Dalhousie, 2712; Liverpool Bay, Nichol- son Isl., 2834; Great Bear Lake: North shore of Dease Arm, low tundra, 4884; Keith Arm, Russel Bay, 3421; Cape McDonnel, 5101; Mackenzie Distr.: South of Back’s River, 64° 30’ N., 104° 20’ W., July 28, 1936, J. Carroll 27. The above are all in the National Her- barium of Canada. In addition the writer has seen a specimen in the Gray Herbarium, collected by the United States North Pacific Exploring Expedition in 1853-56, labelled “Elyna caricina Mert. & Koch, Arakamtchetchene Island, Behring Straits, C. Wright coll.” This appears to match our material, although an uninitialed note on the sheet, presumably by Mackenzie, says: “Kobresia caricina.” Kobresia arctica is perhaps most closely related to K. macrocarpa Clokey, of alpine peaks in Colorado, which, however, is well-marked from ours in having “the terminal flowers staminate and the lateral pistillate’ and by having leaves that equal or exceed the culm. Kobresia schoenoides of alpine Eastern Asia is a much more robust plant with culms 35 to 50 cm. tall and with very conspicuous and shiny old sheaths. One sheet at Copenhagen thus named, and labelled “PI. lenensis subarcticae, Sib. Jakutsk, Bulkur 72°, Sept. 1898, Herm. Nilsson,’ is much taller than our plant but otherwise seems closer to K. arctica than to K. schoenoides. Carex nardina Fries var. Hepburnii (Boott) Kiikenth. in Pflanzenr. 38(1V. 20): 70. 1909. Carex Hepburnii Boott in Hook. FI. Bor.-Am. 2: 209, tab. 207. 1839; Mackenzie in N. Am. Flora 18: 22. 1931, tab. 2. 1940. Boott’s name was based upon a plant from the Canadian Rocky Mts. and was published a few months after Fries had described his C. nardina, based upon material from Greenland. Boott, however, had seen specimens of the Green- land plant sent to him by Hornemann, for under the distribution of C. Hepburnii he added “Greenland, Horn. in herb. Hooker.” In the National Herbarium of Canada there is a duplicate from Boott’s herbarium labelled “C. nardina Fries—C. Hepburnii Boott—Greenland.” Boott, nevertheless, could not have 16 SARGENTIA \4 known the Greenland plant very well, for he does not compare the two and his description of C. Hepburnii is in such general terms that it applies almost equally well to C. nardina. The details of his plate, however, leave no doubt as to the identity of his C. Hepburnii, whereas the drawing of the plant itself is not charac- teristic of that species. It is easy to understand, therefore, that the critical stu- dent of Carex, S. Drejer (Revis. Crit. Caric. 437. 1841) reduced C. Hepburnii to synonymy, although it is worth noting that in the discussion he remarks “Figura cl. Boottii bene habitum nostrae planta C. nardina exprimit, sed de- scriptio perigynii et analyses depictae non satis correctae.’” Steudel (Syn. PI. Cyp. 184. 1855) followed Drejer, while much later Kukenthal (Pflanzenreich 38 (IV. 20): 70. 1909) took up Boott’s name as a variety of C. nardina but reserved it for the plant of the mountains of Pacific N. W. America. At the same time he gave a brief but good description of the plant. Rydberg (FI. Rocky Mts. 119. 1917), giving as a synonym C. stantonensis Jones, and others have applied C. Hepburnii to the plant of western America, excluding C. nardina altogether. Here, then, the matter rested until Tengwall (Sv. Bot. Tidskr. 10: 543. 1916), in reporting C. Hepburnii from northern Scandinavia, started a lot of trouble. Ostenfeld (Medd. om Grgnl. 54: 164. 1923), having carefully examined a large number of plants of what had heretofore been considered good C. nardina from arctic Europe and Greenland, decided that Tengwall was essentially right and that a number of plants from Lapland, Spitsbergen, Greenland, and N. E. Amer- ica should be separated from C. nardina; but he concludes: “I feel it impossible to keep the two forms as distinct species.” This is not surprising, since neither Tengwall nor Ostenfeld had seen specimens of true C. Hepburnii, and the plant they had in mind and which a number of subsequent writers have accepted as C. Hepburnii is actually C. nardina var. atriceps Kukenth. (in Rep. Sp. Nov. 8: 7. 1910), based upon specimens from West Greenland; of this Kukenthal, 1. c., says: “Diese Varietat steht zwischen der typischen Form und der nordameri- kanischen var. Hepburnii (Boott) Kiikenth.” This very clearly is Tengwall’s plant, of which the writer has a duplicate from the original collection and which Tengwall illustrates, |. c., fig. 4 (left). That C. “Hepburnii”’ of Tengwall and Ostenfeld cannot be maintained as spe- cifically distinct from C. nardina has been made abundantly clear by Scholander (Skr. om Svalbard, No. 56: 121. 1933, and 62: 58-62. 1934), who even goes a step farther and, perhaps prejudiced by his personal experiences with the “false” C. Hepburnii, questions the existence of a ‘‘real” one. Having seen no authentic material of C. Hepburnii, he is not convinced by Mackenzie’s treatment, 1. c¢., and, after citing Mackenzie’s key, concludes: “characteristics here enumerated [seem] to be extremely vague and in both instances very variable tendencies which are without any firm inter-coupling.” True C. Hepburnii is poorly represented in most herbaria; even in the Gray Herbarium there are but a few sheets. Therefore, in the absence of authentic material, it is easy to understand why so many recent writers, including Polunin, l. c., although he admits to having seen Scholander’s treatment, have been mis- led by Tengwall and Ostenfeld. The two varieties of C. nardina may be distin- guished thus: C. nardina and var. atriceps var. Hepburnii spike : ovoid-orbicular ovate-oblong staminate part: not conspicuous conspicuous pistillate scales: dark, with pale midvein only conspicuous light center perigynia : distinctly short-stipitate tapering below, not conspicuously stipitate sheaths : gradually tapering into blade abruptly contracted into blade 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 17 Distribution: Carex nardina and its var. atriceps: Extreme northern Europe, Spitsbergen, Iceland, E. and W. Greenland, eastern North America from Ellesmere Island south to Gaspé, Que., west to Hudson Bay and to mountains of western America; not in the Macken- zie District and probably not in Alaska. Carex nardina var. Hepburnii: Mountains of Alberta, British Columbia, and Washington, north to Alaska and ? Yukon; doubtful in Colorado. Unfortunately the matter is still further complicated by the presence of still another species which has passed as C. Hepburnii but which is probably not even very closely related to it. This is the following species. Carex elynaeformis sp. nov. Dense caespitosa, rhizomate brevi ascendente; culmi erecti, tenues strictique, 15-20 cm. alti, obscure triangulares, valde canaliculati, glabri, folia aequantes aut parum superantes, pallide virides; vaginae annorum praecedentium diu_per- sistentes filamentosiusculae, haud nitidae ; folia numerosissima, setiformia, angus- tissima, stricta, canaliculata perconvoluta quapropter facie teretia, marginibus scaberrimis; spica ‘solitaria, ebracteata, 1-1.3 cm. longa, androgyna interdum tota mascula, parte mascula valde conspicua linearisque, quam spica fere dimidio minor; utriculi 5 vel 6 adpressi-ascendentes ; squamae orbiculares, obscure brun- neae, margine pallido membranaceo, utriculi longitudine aequantes; utriculus plano-convexus vel compresse trigonus, 44.6 mm. longus, 1.3-1.5 mm. latus, stramineus apice obscurus, oblanceolatus, inferne in stipitem crassam desinens in rostrum brevem membranaceum contractam, marginibus superne scabris; nux facie lanceolata, valde trigona, circa 3 mm. longa; stigmatibus 3 (2). Great Bear Lake: North shore of Smith Arm, Olmsted Bay, about 66° 32’ N., 122° 30’ W., sandy cutbanks, July 16-21, 1928, 5057 (type). A good match for the above is a plant in the National Herbarium of Canada, represented by three sheets: Cotorapo: Lake Creek, 12,000 ft., July 30, 1919, 7. W. Clokey 3403; Mt. Elbert, 12,000 ft., July 20, 1919, J. W. Clokey 3402; Colorado Flora No. 4022, dry hills, Fall River Pass, Aug. 11, 1921, J. W. Clokey et al. 3485, (all distributed as C. Hepburnii). Carex clynaeformis, by its prominent staminate part of the spike, approaches C. filifolia Nutt., but the scales are not papery white. From C. nardina and its var. Hepburnii it is abundantly distinct by its pale, conspicuously trigonous lanceolate achenes. In all C. nardina and its varieties the achenes are lenticular and of a very characteristic bluish gray colour. Carex elynaeformis appears to be another of the group of bicentric plants known only from high peaks of Colo- rado and from the northern Mackenzie District. Carex chordorrhiza Ehrh. MACKENZIE River Detta: Richards Isl., 2182; Kittigazuit, 2436; Arctic Coast: Atkin- son Pt., 2576; Liverpool Bay, 2849; Eskimo LaKe Bastin: 2nd lake, 2972; Setidgi Lake, 3102; Great Bear Laxe: North shore, 5029; Smith Arm, 5073; McTavish Arm, 5260; KEEWATIN Distr.: Lake on Tha-anne R., 60° 58’ N., 97° W., 5559; west coast of Hudson Bay, 62° 5’ N., 5629; Chesterfield, 6143; Kazan R., Yathkyed Lake, 5761. Not uncommon from the Mackenzie Delta east to Hudson Bay, and north to a short distance beyond the tree limit. Not previously recorded from the Mac- kenzie District. Carex diandra Schrank. Great Bear Lake: McTavish Arm, 3702. Not previously known in the district from north of Great Slave Lake. Carex disperma Dewey. MACKENZIE River Detta: Aklavik, 7303; Great Bear LAKE: McTavish Arm, 5339. New to the flora of the Northwest Territories. 18 SARGENTIA \4 Carex tenuiflora Wahlenb. MackenzigE River Detta: Aklavik, 730¢; Eskimo Lake Basin: Setidgi Lake, 3132; GREAT Bear LAKE: McTavish Arm, 3619, 5262, 5343. Not previously known in the district from north of Great Slave Lake. Carex norvegica Willd. Eskimo LAKE Basin: East of Mackenzie Delta, 2964. New to the flora of the Mackenzie District. Carex neurochlaena Ilolm, in Am. Jour. Sci. 166: 453. 1903, 167: 301, 303. f..1, 2. 1904. MackeENzIeE River DeEtta: Kittigazuit Is!., margin of summer-dry creek, 2337; Arctic Coast: 6 mi. east of Kittigazuit, rare in wet tundra, 2439; Atkinson Pt., in dry Care«-bog some distance from coast, 2564; Eskimo LAKE Basin: North shore of 2nd lake, common in Sphagnum-bog, 2970; Liverpool Bay, in a Carex-bog, 2851; Great BEAR Lake: North shore, 5031; Cape McDonnel, 5775. Somewhat rare or occasional, in dry tundra from the Mac- kenzie Delta east to Anderson River, south to Great Bear Lake. Mackenzie (N. Am. Flora 18: 90. 1931) has reduced C. neurochlaena to syn- onymy under C. marina Dewey (C. glareosa var. amphigena Fern.). With the abundant material now available this seems entirely unjustified. Carex glareosa as well as var. amphigena are true sea-shore plants, while C. neuwrochlaena is an inland plant. In the former two the culm is weak and generally decumbent, while in C. neurochlaena it is slightly curved, firm, and erect. Also in the latter the beak is slightly, but very characteristically, bent. Carex neurochlaena was previously known only from the type locality: Above Rink Rapids, Yukon R., J. Macoun 53879 (Can). The type is annotated by Mackenzie: “C. heleonastes.”’ Carex bonanzensis Britt. This rare Carex, previously known from a few collections near the type locality (Bonanza Cr., Dawson, Yukon) and from FE. Siberia, was found growing abun- dantly in wet places in a meadow in the Mackenzie River Delta, East Branch, 68° 40’ N., Aug. 6, 1933, 6879. New to the flora of the Northwest Territories. Carex leptalea Wahlenb. GreEAT Bear Lake: McTavish Arm, east shore, 3762; southeast shore, 3618. Carex obtusata Liljebl. Arctic Coast oF YUKON Territory: 7124; RicHarpson Mrs.: West of Mackenzie Delta, 6641; MACKENZIE River Detta: East Branch, 6890, 7015; Richards Isl., 2060; Kittigazuit Isl., 2330; 6 miles east of Kittigazuit, 2450, 2451; Arctic Coast: Tuktuayaktoq, 7408; Liverpool Bay, 2848; Eskimo Lake Basin: Kugaruk, 3078; Great Bear LAKE: North shore, 4926, 4927; Dease Arm, 4830. Common or occasional in sandy places east to Great Bear Lake. New to the flora of the Northwest Territories. Carex deflexa Hornem. GREAT Bear LAKE: Dease River Valley, 4870; McTavish Arm, southeast shore, 3644; KEEWATIN Distr.: Tha-anne R., 60° 58’ N., 97° W., 5560. New to the flora of the Northwest Territories. Carex concinna R. Br. MACKENZIE River Detta: East Branch, 6963; Campbell Lake, 7891; Richards Isl., 2058; Eskimo LAKE Basin: Setidgi L., 3737; Great Bear LAKE: Ft. Franklin, 3/97, 3210; Dease Arm, 4755, Carex rupestris All. MaAcKENZIE River Detta: Campbell Lake, 1892; Richards Isl., 2/78; Arcric Coast: Atkinson Pt., 2582; Cape Dalhousie, 2714; Great BEAR LAKE: Cape McDonnel, 51/17. 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 19 An eastern type, rare or occasional in the Mackenzie District, and probably reaching its western limit near the Mackenzie River. Carex glacialis Mackenzie. MACKENZIE River Detta: East Branch, 7016; Campbell Lake, 1890; Great BEAR LAKE: Dease Bay, 4732, 4793; McTavish Bay, southeast shore, 3645; Great Bear River, Mt. Charles, 3290; KEEWATIN Distr.: Kazan R., Yathkyed L., 5766; Tha-anne R., 5564. Rare or occasional in sandy and gravelly places. Carex eburnea Boott. MacKENZIE River De_ta: Campbell Lake, 1894; Great Bear Lake: Great Bear River, Mt. Charles, 3289. Previously known in the Mackenzie District only from the type locality, Ft. Norman. Carex rufina Drejer. Typical specimens of this rare species, previously known only from Greenland and northern Norway, were collected in Keewatin District: Near the edge of a summer-dry pond on Tha-anne R., 60° 58’ N., 97° W., July 12-13, 1930, 5566. Mackenzie (N. Am. Flora 18: 232. 1935) gives the distribution of C. rufina as: “High northern sections of Greenland, etc.” This is erroneous, because C. rufina is not an arctic species. On the west coast of Greenland it reaches 69° 15’ N., but in most years at its northern limit it is unable to mature fruits. Carex rufina, although it forms pure stands, is a most inconspicuous species, easily overlooked because normally the flowering culms are completely concealed among the leaves. Carex bicolor All. Arctic Coast: 6 miles east of Kittigazuit, 2446, 2447; Cape Dalhousie, 2716; Great BEAR Lake: Dease Arm, north shore, 4889; Haldane R., 4918, 4919; Smith Arm, north shore, 5058; Krewatin Distr.: Lake on Tha-anne R., 5557; Mistake Bay, 5628. Not previously recorded from the Mackenzie District. Carex Garberi Fern. MaAcKENZIE River Detta: South end of Richards Isl., 7062; north end, 2183; Great BEar Lake: Great Bear River, 3264; Dease Arm, 4792; McTavish Arm, 3675, 5288, 53235. Not previously recorded from north of Great Slave Lake. Carex livida Willd. var. Grayana (Dewey) Fern. Eskimo Lake Basin: 2nd lake, 69° N., 132° 30’ W., 2965. Our specimens grew in a large stand in a wet meadow, and on Aug. 18, 1927, had mature fruit. New to the flora of the Northwest Territories and not before recorded from the Mackenzie basin. Carex Williamsii Britt. MackeEnziE River Detta: North end of Richards Isl., 2773; Arctic Coast: 6 miles east of Kittigazuit, 2448; Liverpool Bay, 2845; Eskimo Lake Basin: 2nd lake, 2968; KEEWATIN Distr.: Kazan R., Yathkyed Lake, 5759. New to the flora of the Mackenzie basin. Carex Oederi Retz. var. pumila (Cosson & Germ.) Fern. GreaT Bear LAKE: South shore, Leith Pt., 3530, 3558; Cape McDonnel, 5126. Not previously recorded in the Mackenzie country from north of Great Slave Lake. Carex petricosa Dewey. RicHArpson Mrs.: Eastern slope, west of Mackenzie Delta, alpine tundra, 2000-3000 ft., in beginning anthesis, 6647, 6648; Arctic Coast: Liverpool Bay, 70° N., 129° W., dry edge 20 SARGENTIA \4 of a small tundra pond, fruiting specimens, Aug. 15, 16, 1927, 2840; Great Bear LAKE: North shore of Dease Arm, 67° 02’ N., 119° 50’ W., flowering specimens, June 23-27, 1928, 41676, 4887. This very striking species has been collected but a few times, all in the general vicinity of the type locality in the Canadian Rockies: Banff, July 14, 1900, N. B. Sanson 25498 (Can); Mt. Rendle, 6000’, Aug. 7, 1902, N. B. Sanson (G); Banff, Aug. 4, 1891, J. Macoun 7425 (Can). Our specimens are a close match for these. In some of the spikelets of our no. 2840, a few perigynia have two instead of three stigmas. Carex atrofusca Schk. Common in not too dry tundra from mountains west of Mackenzie Delta to Hudson Bay, south to Great Bear Lake, where it is alpine or restricted to “bar- rens” and absent in wooded parts. As the writer has pointed out (in Rhodora 41: 204. 1939), it is strange that this species, common throughout the “barren grounds,” was not collected by Richardson, and that it has not previously been collected in the Mackenzie District. Carex atrofusca Schk. var. decolorata var. nov. A specie differt squamis anguste lanceolatis, cinnamomeis, nervo medio ob- scuro, utriculisque cinnamomeis basi nigro-purpurascentibus, 3.5—4+ mm. longis ; nuce 1.4 mm. longa. Great Bear Lake: Cape McDonnel, low tundra, Aug. 2, 1928, 5120 (tTyPE). Carex rariflora (Wahlenb.) Sm. var. androgyna var. nov. A specie differt spiculis terminalibus androgynis stigmatibus 2 nec 3. Arctic Coast: Atkinson Pt., 70° N., 131° 20’ W., Aug. 1-3, 1927, 2579 (TYPE). Carex limosa L. Eskimo Lake Basin: Setidgi Lake, 3735; Great Bear LAKE: McTavish Arm, north shore, 5220; southeast shore, 3601. In the Mackenzie District not recorded from north of Great Slave Lake. Carex paupercula Michx. Great Bear Lake: McTavish Arm, north shore, 5221; east shore, 3723. Not previously recorded in the Mackenzie District from north of Great Slave Lake. Mackenzie (N. Am. Flora 18: 351. 1935) points out that C. paupercula Michx. is not to be confused with the South American C. magellanica Lam., which has “the terminal spike normally gynaecandrous.” In our no. 5227 fully half of the terminal spikes are gynaecandrous. Carex media R. Br. in Richardson in Franklin’s Journ, App. 750. 1823. Carex Vahlti Schk. var. inferalpina sensu Fernald in Rhodora 35: 220-223, 398. 1935, non Wahlenb. Carex angarae Steud. See Rhodora 41: 203-205. 1939. Common in the Mackenzie District, in the wooded parts, north to the Mac- kenzie Delta (Eskimo Lake Basin, Setidgi Lake, 3738) and Great Bear Lake. Carex norvegica Retz. Fl. Scand. Prodr. ed. 1. 179. 1779, non Willd. Carex Vahlit Schk. KEEWATIN Distr.: West coast of Hudson Bay west to Kazan R., Yathkyed Lake, 5757, 6142. An eastern type not known west of the last-mentioned station. Carex holostoma Drejer. MacKenzie River Detta: Richards Isl., 2/81, 7064; Arcric Coast: 6 miles east of Kittigazuit, 2442; Cape Dalhousie, 2723; Liverpool Bay, 2846; Great BEAR LAKE: McTavish 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 21 Arm, east shore, 3711, 3722; north shore, 5259; fjords on the east shore, 5342; KEEWATIN Distr.: Kazan R., Yathkyed Lake, 5762; Baker Lake, 6080; Lake on Tha-anne R., 5561; Mistake Bay, 5630; Chesterfield, 6144. Although Polunin (1. c., 129) records this species as “new to North America, outside Greenland,” the writer and his brother thirteen years earlier, in 1927, collected it from numerous stations in the Mackenzie District. Carex holostoma is a truly arctic species widely distributed in arctic parts of North America, from the Mackenzie R. east to central parts of West Greenland, south to the edge of the forest. It is absent in East Greenland and Iceland but has an isolated occur- rence in northernmost Scandinavia. The illustration in Mackenzie, |. c., tab. 474, 1940, is very poor; a much better one is in Ostenfeld, Fl. Arct. fig. 32. 1902. Carex stylosa C. A. Mey. Great Bear Lake: Cape McDonnel, 5105/4. Not previously recorded from the Northwest Territories. Carex podocarpa R. Br. See Rhodora 41: 205-206. 1939. Ricwarpson Mrs.: 6645, 6646, 6763, 6764, 6766, 6767; Arctic Coast: West of Mackenzie Delta, 6903, 7090. Common in mountains west of the Mackenzie Delta, north to the arctic coast, on snow-flushes and in herb mats. Carex nesophila Holm. Ricuarpson Mrts.: Eastern slope, west of Mackenzie Delta, 4000’ elev., 6644. New to the flora of the Northwest Territories. Carex atrosquama Mackenzie, Great Bear Lake: Smith Arm, north shore, 4991; Keith Arm, Deerpass Bay, 3457; Etacho Pt., alpine meadows, 3476. Rare or occasional in meadows in the Paleozoic country about the west end of the lake. New to the flora of the Northwest Territories. Carex albo-nigra Mackenzie. MackENZIE River De_ta: East Branch, 68° 55’ N., 7017; Arctic Coast: East of Mac- kenzie Delta, 25834; Great Bear Lake: Cape McDonnel, 5104; McTavish Arm, north shore, 5289. In rather wet tundra. New to the flora of the Northwest Territories. Carex Buxbaumii Wahlenb. Great Bear Lake: South shore, Leith Pt., 3532, 3566. Typical specimens were found on wet shores of a limestone island. New to the flora of the Northwest Territories. Carex Morrisseyi sp. nov. Rhizomata longa tenuia non-stolonifera; culmi solitarii aphyllopodici tenues erecti, 20-35 cm. alti, acutanguli superne scabri; folia plana, circa 3 mm. lata, marginibus scabris, revolutis, aeque ac culmi cinereo-viridia, culmo dimidio bre- viora; spiculae 3 vel 4 valde approximatae, sessiles vel inferior breviter pedun- culata, mascula (vel rarissime androgyna) terminalis linearis vel lineari-clavata, 10 mm. longa, laterales 2 vel 3 foemineae, ellipsoideo-obtusae, 10-12 mm. longae, densiflorae; bractea inferior setacea quam spicula multo brevior, evaginata; squamae foemineae ovatae vel ovato-cuspidatae (nunquam aristatae) nigrae, nervo medio pallido prominente (squamae masculae longiores angustioresque, obscure castaneae) ; utriculi squamas superantes, ovato-ellipsoide1 subcompressi cinereo- virides superne distincte papillosi, 3 mm. longi, brevissime stipitati, in rostrum brevissimum fauce integro abrupte contracti; stigmata brevia 3 (aliquando 2) ; nux obovata sessilis, trigona, 2 mm. longa, in stylum brevem contracta. 22 SARGENTIA |4 Wet, marshy places on pre-Cambrian rock from Labrador coast to Great Bear Lake. LABRADOR: Cutthroat Harbor, south of Cape Mugford, 57° 30’ N., 62° W., low granite island, Aug. 26, 1937, Bartlett Greenland Expedition, 1937, 4. E. Porsild 173 (tyrE) ; Okkak, near Cutthroat Tickle, 57° 40’ N., 62° W., Aug. 12-14, 1937, V. C. Wynne-Edwards 7523; KEEWATIN Distr.: Yathkyed Lake on the Kazan R., 62° 30’ N., 97°-98° 30’ W., Aug. 1-15, 1930, 5775; Mackenzir Distr.: Great BEAR Lake: McTavish Arm, Conjuror Bay, 65° 50! N., 117° 15’ W., Aug. 15, 1928, 3674; Cape McDonnel, low tundra and lake shore, Aug. 2, 1928, 5105; north shore, 4991A, Carex Morrisseyi commemorates Captain Robert A. Bartlett’s schooner Effie M. Morrissey, veteran of numerous arctic voyages. It belongs in the section Atratac Kunth and admittedly is closely related to C. Burbaumit Wahlenb. (C. polygama auctt., non Gmel., fide Mackenzie), of which polymorphous species var. hetero- stachya Anders. also has the terminal spikelet staminate (“spicula terminalis mere 3”) ; but it is certainly abundantly distinct from all American forms of that species. Carex Buxbaumiu of Europe has recently been revised by Cajander (Ann. Bot. Soc. Zool.-Bot. Fenn. 5: 5. 1935), who divides the species into several sub- species and varieties and who reports that plants with staminate terminal spike- lets are not rare in Fennoscandia. He assumes (under C. polygama ssp. alpina (Hartm.) Cajander) that Hartman “with his y heterostachya by and large had in mind specimens belonging just with this subspecies” (translated from Ca- jander’s German text, |. c. 22). Carex polygama ssp. alpina, however, is described and illustrated with a gynae- candrous terminal spikelet, and, while in Scandinavia staminate terminal spikelets may occur in all “forms” of the collective species, in our plant this is the rule. Thus, at the type locality, in a large colony, but three specimens could be found in which the terminal spikelet was androgynous and none that were gynae- candrous. Also in our plant the spikelets are more densely aggregate and but rarely with the lowermost spikelet somewhat remote; moreover it appears to have a distinct sub-arctic geographical range of its own. Intermediate forms between it and C. Bicrbawmii have never been observed, nor have variations in C. Buxbaumiu ever been recorded from North America having other than “oblong-ovoid gynaecandrous” terminal spikelets. Carex anguillata Drejer in Revis. Crit. Car. 454. 1841; Steudel in Syn. PI. Cyp. 214. 1855; Fl. Dan. 16: tab. 2846, 1871. Great Bear LAKE: Smith Arm, north shore, 4988; Cape McDonnel, 5114, 5109; McTavish Bay, north shore, 5172; Keewatin Distr.: Kazan R., Yathkyed L., 5767, 5774; Tha-anne R., 5565; Chesterfield, 6146. Carex anguillata has been referred by Boott to C. aquatilis var. minor, whereas by most other writers it is placed with C. rigida Good. Both interpretations seem unsatisfactory in view of the long-peduncled, long- and lax-flowered pistil- late spikelets, although there can be little doubt that the affinity is with the latter. Fernald, in Rhodora 36: 91. 1934, drew attention to Drejer’s C. anguillata, almost forgotten since Kukenthal (1. c., 302) reduced it to an unimportant form of C. concolor R. Br., citing specimens from Iceland and from Labrador (4. P. Low 13459, Can). Fernald suggested its presence in Greenland and Maine, while Mackenzie, |. c., cited material also from Hudson Bay and Keewatin. Carex anguillata seems amply distinct from C. concolor and is not rare in S. Greenland, Labrador, and Keewatin, west to Great Bear Lake. It grows in sandy well-drained soil, where it forms a loose turf. The specimens cited above from the writer’s collection are fairly typical. 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 23 Carex concolor R. Br. seems to be absent in continental parts of N. America west.of Hudson Bay. Carex lugens Holm. See Mackenzie in N. Am. Flora 18: 401. 1935, excl. C. consimilis Holm; also, Porsild in Rhodora 41: 208. 1939. Arctic Coast oF YUKON Territory: 7088, 7126; Ricuarpson Mts.: 66424, 6643, 6765; Mackenzie River Detta: East Branch, 6537, 6957, 6990, 6991, 7017A; Campbell Lake, 1893; south end of Richards Isl., 2059, 2064, 2065, 7063; north end of Richards Isl., 2164, 2175-2177 ; Kittigazuit, 2325; Arctic Coast: East of Delta, 2444; Eskimo LAKE BasIN: 2nd lake, 2963. The above series is rather uniform; most numbers have a few staminate flowers at the apex of the lateral spikelets. Carex lugens is a common species in Alaska, Yukon, and Northwest Mackenzie District east to Keewatin. It grows in old, well-established tundra where it forms large, compact tussocks. Carex consimilis Holm in Am. Jour. Sci. 166: 457. 1903, 167: 310. 1904. Besides the type the following specimens, all in the National Herbarium of Canada, are typical: YuKoN Territory: Lake Kluane to Don Jek R., A. Miiller, Aug. 11-27, 1920; MackENziE De ta: Kittigazuit, 2323; Arctic Coast: East of Mackenzie Delta, 2443; At- kinson Pt., 2583; Cape Dalhousie, 2730; Liverpool Bay, 2853; GreAT BEAR LAKE: North shore, 4922; Cape McDonnel, 5103, 51284; Keewatin Distr.: Kazan R., Yathkyed Lake, a7 as This species has been placed by Mackenzie, |. c., under C. lugens Holm, but the type (John Macoun 53878, Can) differs so much from C. lugens in its strong, horizontal rhizomes, the short and broad leaves and stout culms, etc., that this interpretation is most unsatisfactory. Moreover, there is in the writer’s collection a series of plants that completely match the type as well as the description. Lack- ing the dense, caespitose habit of C. lugens, these plants, in the field, had been tentatively labelled C. concolor for lack of a more satisfactory name, although it was even then realized that true C. concolor is absent in continental parts of North America west of Hudson Bay and that our plant, moreover, differs con- sistently from it by its short, oval-oblong, almost black, contiguous spikelets. Carex consimilis grows in dry peaty places in lowland tundra. Carex aquatilis Wahlenb. Common in the Mackenzie valley north to the Arctic Ocean and east to Great Bear Lake, chiefly in alluvial soils along the north-flowing rivers. It is a truly boreal, though not at all an arctic species, which in the delta of the Mackenzie reaches slightly beyond the limit of trees. Along the Mackenzie it forms ex- tensive meadows and in the delta it is one of the principal sources of native hay. Carex stans Drejer, Revis. Crit. Car. 458. 1841; Fl. Dan. 14: tab. 2477 (poor). 1849; Steudel, Syn. Pl. Cyp. 210. 1855; Lange, Consp. Fl. Groenl. 147. 1880. Carex aquatilis var. stans (Drejer) Ostenf. Fl. Arct. 70, fig. 43. 1902. Carex aquatilis var. minor Boott, Carex 4: 163, tab. 545 (good). 1867. Carex stans differs from C. aquatilis in its shorter and stouter culms, which are nearly always longer than the leaves; the latter are flat, or the margins may be slightly involute, generally broader, never conduplicate or glaucous as in C. aquatilis. The sheath of the lower bract is more loosely clasping and generally black-auricled. The staminate spikelet is solitary (rarely 2); the pistillate ones are shorter and thicker than in C. aquatilis. Perigynia tend to be slightly longer, stramineous rather than green, and often dark-tipped, smooth, under a lens not conspicuously punctate; scales somewhat longer, often equaling the perigynia, dark brown or even black with a pale mid-vein (much darker than in C. aquatilis ). 24 SARGENTIA \4 Carex stans in its distribution is truly arctic. That it is not merely an arctic form or race seems amply proved by the fact that in W. Greenland it reaches south only to lat. 68° and is absent on the east coast south of 76°. South of these latitudes it is not replaced by C. aquatilis, which appears to be wanting in Greenland. Polunin, 1. c., 132, records C. aquatilis from the eastern Canadian Arctic, but mentions “intermediate” forms between that species and “var. stans.” Polunin probably did not understand C. aquatilis, and all records of that species from the eastern Arctic should be changed to C. stans. Distribution: Arctic-circumpolar (except Iceland); across N. America, and southward almost to the limit of trees. Carex salina Wahlenb. MACKENZIE River Decta: North end of Richards Isl., 2166, 7448; Eskimo LAKE Basin: Kugaruk, 3080. Our plant is intermediate between typical C. salina Wahlenb. and C. sub- spathacea Wormskj., but it is a much coarser plant than the latter. It is C. salina a stricta Drejer (in Revis. Crit. Car. Bor. 452. 1841), which Ktkenthal reports from the Bering Sea (Macoun 16619), and probably also C. Hopnerii Boott (in Hook. Fl. Bor.-Am. 2: 219. tab, 220. 1839). Also it is a near match for specimens from the Bering Sea referred by Hultén to C. Ramenskii Kom. (Rep. Sp. Nov. 13: 164. 1914). Typical C. subspathacea Wormskj., in America, seems to be an eastern arctic species, on the arctic coast reaching west to Dolphin and Union Strait (Can. Arct. Exp. 341, 378, Can). Carex microglochin Wahlenb. GREAT Bear Lake: North shore, Haldane R., 5032; Cape McDonnel, 5127; MeTavish Arm, north shore, 5171, Not previously recorded from the Mackenzie District. Carex physocarpa Presl. MACKENZIE River Detta: East Branch, 6989; Aklavik, 3183; Kittigazuit Isl., 2322; Arctic Coast: 6 miles east of Kittigazuit, 2447; Atkinson Pt., 2570, 2571; Cape Dalhousie, 2724; Esximo Lake Basin: 2nd lake, 2967. A western arctic species common in Alaska and Yukon, eastward to the Mac- kenzie Delta, and in the alpine Rocky Mountains to the south. All records from east of the Mackenzie River and east of the Rocky Mountains no doubt should be referred to C. saxatilis L. Compare also Polunin, 1. c., 136. Carex saxatilis L. An eastern arctic species, the N. American distribution of which includes E. & W. Greenland, N. Labrador, Baffin Isl., and the country from Hudson Bay west to Lake Athabaska and Great Bear Lake. Carex membranacea Hook. One of the most common and most truly arctic sedges, abundant from Alaska to northern Labrador and south to the limit of trees. It is strange that it has not reached Greenland. Carex rotundata Wahlenb.; Hook. Fl. Bor.-Am. 2: 220. 1839, Bear Lake record only; Mackenzie, N. Am. Flora tab. 518, 1940, not as to text and distribution (18: 454. 1935); Gray’s Man. ed. 7. 255, fig. 552, excl. distr. MackenziE River De_ta: North end of Richards Isl., 2163; Eskimo Lake Basin: North shore, 2nd lake, 2971; Great Bear LAKE: ? Fort Enterprise, Richardson. 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 25 Mackenzie, |. c., appears to have been misled by writers on the flora of Green- land, including Ostenfeld in Fl. Arctica, who have confused C. rotundata with C. saxatilis. Why this should be is not easy to understand, because C. rotundata is a clear-cut and well-marked species. It is well-described and illustrated in Gray’s Man., |. c., from a typical specimen in the Gray Herbarium labelled: “C. rotundata Wahl. dedit Wahl. from I. Br. herb.” Mackenzie’s plate also is very good. Carex rotundata Wahlenb. is an arctic Eurasian species which enters N. Amer- ica in Alaska and northwest Mackenzie. It is a very rare plant in American herbaria, where most plants so-named should be referred to C. savratilis or to C. membranacea. Although for a hundred years it has been recorded from Greenland by most authors on that flora, the present writer has never, in many years of collecting in that country, been able to satisfy himself as to the authen- ticity of Greenland material thus named. Recently Sgrensen (Medd. om Gr@gnl. 101°: 122. 1933, 101*: 83. 1934) has voiced such doubt, stating simply that, by comparison with authentic Scandinavian material, all Greenland specimens previ- ously called C. rotundata are referable to C. saratilis. With this view the writer is in complete accord. Carex oligosperma Michx. Although Richardson collected this species on Great Bear Lake, the writer did not see it there. It was collected recently on the Yellowknife R., 63° 30’ N., au V,, by J. Carroll (8, Can). LEMNACEAE Lemna trisulca L. ; Mackenzie River Detta: East Branch, 68° 40’ N., 7375; Aklavik, 68° 12’ N., 135° W., flowering sparingly in a shallow lake, Sept. 1, 1934, 7305; lake between delta and Eskimo Lakes, 68° N. (fragments in muskrat “push-up” on the ice), Nov. 2, 1933, 6932. Lemna trisulca in the Mackenzie basin was previously known north to Lake Athabaska and the Wood Buffalo Park, 60° N. (Raup), and the above numbers, therefore, represent a considerable extension of the known range. ARACEAE Calla palustris L. MACKENZIE River Detta: Lake on the portage trail between Campbell and Setidgi Lakes, about 68° N., 133° W., rhizomes with young leaves, June 16, 17, 1927, 2023; Eskimo LAKE BaAsIN: Creek above Setidgi L., plants with mature fruit, Aug. 24, 1927, 3175; tundra lake between delta and Eskimo Lakes, fragments from muskrat “push-up,” Nov. 2, 1933, 6933; MACKENZIE River: Lake near Sans Sault Rapids, 65° 40’ N., Sept. 20, 1928, 3400. Calla palustris was previously known north to Great Slave Lake (Raup). The above numbers represent a considerable extension of the known range. LILIACEAE Tofieldia nutans Willd. See Porsild in Rhodora 41: 213. 1939. MackKENZIE River Detta: Husky R., in a muskeg, 6779. New to the flora of Canada. Lloydia serotina (L.) Reichb. RicHarpson Mrs.: West of Mackenzie Delta, rare in alpine tundra, 6653; Arctic Coast oF YuKON Territory: King Pt., 7130. Alaska-Yukon east to Mackenzie R. New to the flora of the Northwest Territories. 26 SARGENTIA [4 Smilacina stellata (L.) Desf. Great Bear LAKE: Great Bear R., Foot of Mt. Charles, 33/0. New to the flora of the Northwest Territories. ORCHIDACEAE Cypripedium parviflorum Salisb. MACKENZIE River: Bear Rock at Ft. Norman, 3381; Great BEAR LAKE: Great Bear River, Foot of Mt. Charles, 3291, 3326; west shore, 65° 30’ N., J. M/. Bell 22974 (Can). New to the flora of the Northwest Territories. Cypripedium guttatum Sw. MACKENZIE River: Ft. Norman, G. Fume 103449 (Can). Not previously recorded in our region from north of Great Slave Lake (Raup). Orchis rotundifolia Pursh. MacKENZIE River DeLtta: Campbell Lake, 1983; Great Bear LAKE: MeTavish Arm, east shore, 3622, 3763; southeast shore, 5296; Leith Pt., south shore, 3576. Not previously recorded from north of Great Slave Lake. Habenaria obtusata (Pursh) Richards. Common in the Mackenzie region north to the limit of trees, occasional or rare for a short distance beyond. The known north limit is on the Arctic Coast: 6 miles east of Kittigazuit, 2467. Goodyera repens (L.) R. Br. MACKENZIE River: Sans Sault Rapid, 656/. Previously recorded from Great Bear Lake by Richardson. Listera borealis Morong. MACKENZIE River Detta: East Branch, 6576, 7280; Great BEAR LAKE: Keith Arm, Ft. Franklin; south shore, Leith Pt., 3539; Bear R., 3361. Previously known on the Mackenzie northward only to Ft. Simpson. Corallorrhiza trifida Chat. MAcKENZIE River De_ta: Husky R., 6654; Aklavik, 6575; East Branch, 6540; Campbell Lake, 1984, 1985; Richards Isl., 2190; Arctic Coast: 6 miles east of Kittigazuit, 2460; GREAT BEAR LAKE: Bear R., 3275; McTavish Arm, 5227, 5294, 5294A, 5327; Dease Arm, 4681, 4733, 4892; Keith Arm, Russel Bay, 3409; Etacho Pt., 3486; south shore, Leith Pt., 3537, 3575; MACKENZIE RIvER: Root R., 62° 30’ N., 6273; Kreewatin Distr.: Mistake Bay, 62° 05’ N., 5648. Fairly common in the Mackenzie District, north to the limit of trees or a short distance beyond. Not previously recorded from north of Great Slave Lake or from Keewatin District. SALICACEAE When pushing his way for endless, weary miles through unmapped, pathless “bushland” of the Northwest Territories, in winter by dog team or on snowshoes, in summer on foot or by canoe, the writer was always reminded of the dictum of that great student of North American willows, who said: “We know, how- ever, almost nothing of the Salix flora of the woodland region of the Northwest Territories, which must be an Eldorado for willows” (C. Schneider in Bot. Gaz. 66: 340. 1918). The writer now regrets that before going into the North he was unable to prepare himself adequately for the study of the willows in the field. Although in the Northwest Territories he collected a total of 35 species, represented by 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 27 235 collections aggregating about 1000 sheets, he realizes that too often the col- lecting of critical species was done in a haphazard way and without sufficient notes. The willow collection has been named and studied by the Swedish sali- cologist, Dr. B. Floderus. Pending the publication of Dr. Floderus’ notes the following distributional data on the more interesting species are given. Salix arbusculoides Anders. One of the most common willows of the lower Mackenzie valley, represented by several series of collections. In the Mackenzie Delta, north of the continuous spruce forest, it forms dense thickets and “willow islands” 10-15 feet high, and in the forest it often forms the principal “‘underbrush” with tree-like bushes 15-20 ft. high. North of Great Bear Lake and in the Keewatin District it reaches a short distance beyond the tree limit, where it was collected along Kazan R. at Yathkyed L., 5789. Salix arbutifolia Pall. Salix fuscescens Anders. Arctic Coast oF YuKoN Territory: Shingle Pt., 7093; MACKENZIE River Detta: East Branch, 6955; Arctic Coast: East of Mackenzie, 2467, 2589, 2590, 2866, 2976; KEEWATIN Distr.: Kazan R., 6025. New to the flora of Canada. Salix arctophila Cockerell. A common species in wet tundra north of the forest. Although apparently an eastern species, it extends across the continental Northwest Territories from Hudson Bay to the Arctic Coast west of the Mackenzie Delta (7091, 7172). Salix candida Fliigge. Great Bear Lake: Bear R., 3251. Although it was recorded by Richardson as occurring “throughout the wooded country,” the writer saw this species but once. Salix Chamissonis Anders. Ricuarpson Mrs.: Grassy slopes in the foothills, 1500’, 6781, 7329. This rare willow, previously known only from the Bering Sea region, is new to the flora of Canada. Salix glauca L. var. acutifolia (Hook.) Schneid. in Bot. Gaz. 66: 327. 1918. Salix Seemannii Rydb. MackeNnziE River Detta: Limestone scree, Richardson Mts., 6783; East Branch, 7250, 7253; Campbell Lake, 1904; Arctic Coast of YuKOoN Territory: King Pt., 7131; GREAT Bear‘Laxe: Keith Arm, 3410, 3411; McTavish Arm, north shore, 5263. This is a common and most variable shrub, in our region chiefly on Paleozoic formations or at least with a preference for calcareous soil. Salix fullertonensis Schneid. This eastern species, previously recorded westward only as far as Coronation Gulf, proved very common on “barrens” along the shores of Great Bear Lake, 4685, 4686, 4687, 5135, 5136, 5138A, 5653. Salix glacialis Anders. See Schneider in Bot. Gaz. 67: 63-64. 1919. This rare willow, known only from a few places in Alaska, was collected on a sandy beach near Shingle Pt., Y. T., 7094. New to the flora of Canada. Salix herbacea L. Great Bear LAKke: Dease Valley, 4874; Keewatin Distr.: Kazan R., Yathkyed L., 5787, Lake on Tha-anne R., 5560A ; west coast of Hudson Bay, Mistake Bay, 5650. 28 SARGENTIA \4 This eastern species, which has not previously been collected west of Hudson Bay, is within its range ubiquitous in herb-mats and on snow-flushes. Its scar- city in the interior may be due in part to the aridity of the climate and consequent lack of suitable stations. Salix hudsonensis Schneid. An eastern species, with a distribution pattern somewhat similar to that of S. fullertonensis, collected by us on Great Bear Lake, 5138B, and on the Arctic Coast west to the Mackenzie Delta (numerous collections). Previously recorded by Schneider westward to Coronation Gulf. Salix lingulata Anders. See Schneider in Bot. Gaz. 66: 353. 1918. Arctic Coast oF Yukon Territory: King Pt., 7173; MACKENZIE River De.ta: East Branch, 7366; Campbell Lake, 1986; Kittigazuit, 2468 (X niphoclada) ; Great BEAR LAKE: North shore, 4688A and B, 4689. Salix niphoclada Rydb. See Schneider in Bot. Gaz. 66: 339. 1918. Arctic Coast: Liverpool Bay, 2861, 2862; Great Bear LAKP: Cape McDonnel, 5137; Great Bear R., Mt. Charles, 3327. This little-known willow, the type of which came from the lower Mackenzie River, appears to be fairly common in the Paleozoic country of the northwestern Mackenzie District. Salix nivalis Hook. var. saximontana Rydb. KEEWATIN Distr.: Lake on Tha-anne R., 5574. According to Floderus (in litt.) this is represented in our collections by this single number. Salix pedicellaris Pursh var. hypoglauca Fern. KeEwatTin Drstr.: Lake on Tha-anne R., in a Carex bog, 5573. Represented in the collection by this single number. New to the flora of Keewatin. Salix phlebophylla Anders. Vaccinium caespitosum sensu Holm & Macoun in Rept. Can. Arct. Exp. 5A: 18. 1921, not Michx. This tiny Alaskan willow is rare on dry tundra ridges in the Caribou Hills east of the Mackenzie Delta: East Branch, 7263; Kittigazuit, 2469. Also there is a single collection from Richardson Mts., west of Mackenzie Delta, O. Bryant 6581. The first and only previous collection from Canada was made at Herschel Island, Y. T., by the Canadian Arctic Expedition, but recorded by Macoun & Holm, |. c., as Vaccinium caespitosum. Salix pulchra Cham. One of the most common willows in the Mackenzie Delta, where it grows in alluvial soil, often with S. arbusculoides. It has tentatively been referred by Floderus (in litt.) to S. pulchra var. yukonensis Schneid. Salix pulchra is com- mon in Alaska and in the Yukon and was recorded by Ostenfeld (in Gjéa Exp. 35) from the Arctic Coast west of the Mackenzie Delta. One collection from Richardson Mts., west of the Mackenzie Delta, 7328, also seems to belong to the species. MYRICACEAE Myrica Gale L. Very common and represented by numerous collections from Great Bear Lake. where it had been collected previously by Richardson. Apparently rare in the 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 29 Mackenzie River Delta (Campbell Lake, 1901) and in Keewatin District (Lake on Tha-anne R., 5526). BETULACEAE Betula microphylla Bunge. MacKENZzIE River Detta: West Branch, O. Bryant 6584; East Branch, 7465; Setidgi Lake, 3143; Great Bear LAKE: McTavish Arm, 5229, 5230. Rare or occasional from the Mackenzie Delta east to Great Bear Lake. In the Mackenzie Delta, B. microphylla is a small tree rarely more than 12 feet high, quite unlike B. papyrifera in shape and foliage. The leaves turn fully two weeks earlier than those of the latter. Betula papyrifera Marsh. var. neoalaskana (Sarg.) Raup. According to Dr. Hugh M. Raup, who has kindly examined the birches in the writer’s collection, this appears to be the common tree-like species of the Mac- kenzie Delta, the lower river, and Great Bear Lake. The paper birch is rare in the low parts of the Delta but on the higher east and west banks, and on Bear Lake, it reaches a height of about 30 feet, with trunks 8 inches in diameter. On the East Branch it does not extend quite as far north as the spruce (Picea glauca). Betula papyrifera Marsh. var. occidentalis Sarg. MackKENZIE River: Bear Rock, near Fort Norman. In the Mackenzie Basin outside the mountains not previously recorded from north of Lake Athabaska. POLYGONACEAE Koenigia islandica L. KEEWATIN Distr.: Lower Kazan R., 97° W., 6026 (new western limit). While Polunin, 1. c., states that this plant is circumpolar and Hulten (FI. Kamtchatka 2: 42. 1928) gives its American distribution as “along the Arctic Coast to Baffin Land,” the fact remains that it has never been collected between the Seward Peninsula in Alaska and the Keewatin District. Rumex acetosa L. RIcHARDSON MtTs.: West of Mackenzie Delta, 6656, 6786. New to the flora of the Northwest Territories. Rumex pallidus Bigel. MACKENZIE River Detta: Aklavik, occasional on sandy river banks, 7300; East Branch, 7299, 7307; Richards Isl., 2075; Kittigazuit, 2472. Polygonum alpinum All. var. lapathifolium Cham. & Schlecht. MACKENZIE RIveR Detta: East Branch, 6502, 7295; Peel River, Miss E. Taylor 73 (Can). Our plant is probably “P. alpinwm y foliis ovato-acuminatis glaberrimis. . . . Arctic Sea-coast, Dr. Richardson,” in Hooker, Fl. Bor.-Am. 2: 131, although it actually does not reach north beyond the limit of trees in the Mackenzie Delta. It is taller and more robust than the plant of the Bering Sea region and per- fectly glabrous. It is best developed in ravines on the East Branch, where, on the freshly exposed calcareous clay of recent landslides, it forms pure stands of an acre or more in extent. Here the plant may grow 5 or 6 feet high, with 50 or more stems from one root. The shoots are edible when young and are used locally as rhubarb. 30 SARGENTIA [4 CHENOPODIACEAE Suaeda maritima (L.) Dumort. Chenopodium maritimum L., Hook. Fl. Bor.-Am. 2: 126. 1838. Arctic Coast: Atkinson Pt., 2605; Great Bear Lake: North shore of McTavish Arm, near mineral spring, 5193. PORTULACACEAE Montia lamprosperma Cham. Polunin, |. ¢., suggests that this is cireumboreal, but actually, like Koenigia islandica, it appears to be absent between the western tip of Alaska and the west coast of Hudson Bay. CARYOPHYLLACEAE Arenaria arctica Stev. Apparently common in mountains west of the Mackenzie Delta and along the arctic coast of the Yukon Territory. Richardson’s “Shores of the Arctic Sea,” in Hook. Fl. Bor.-Am. 1: 100, no doubt, refers to the latter place. We never collected it east of the delta nor at Great Bear Lake, although it was collected once in the latter place by J. MW. Bell in 1900 (22887, Can). A still more puz- zling collection is that of A. P. Low, Northern Labrador (Ungava), Aug. 1897 (18258, Can). Arenaria dawsonensis Britt. GreAT Bear Lake: Dease Arm, 4805; north shore, 5078; McTavish Arm, 3779, 5269, 5298. Common on the pre-Cambrian east shore of the lake. Arenaria humifusa Wahlenb. Mackenzie River De_ta: Campbell Lake, 1917; Great Bear Lake: Smith Arm, 4997; McTavish Arm, 3766; south shore, Leith Pt., 3578: Kreewatrn Distr.: Kazan R., 6027; Yathkyed L., 5793. Arenaria laricifolia L. RicHarpson Mrts.: West of Mackenzie Delta, 6797. New to the flora of the Northwest Territories. Arenaria macrocarpa Pursh. , Arctic Coast oF Yukon Territory: King Pt., 71344, 7174, rare or occasional. Thus far not recorded from east of the Mackenzie. Arenaria nardifolia Ledeb. MacKENZIE River Devra: East Branch, 7269; Richards Isl., 2079, 2206; Kittigazuit, 2481; Arctic Coast: Atkinson Pt., 2598; Liverpool Bay, 2872; Great Brar Lake: Dease Arm, 4835; Smith Arm, 4999, 5076; McTavish Arm, 3624. Common in dry places eastward to Great Bear Lake. Arenaria obtusiloba (Rydb.) Fern. RicHarpson Mrs.: West of Mackenzie Delta, 6795; Mackenzie River Devra: Campbell Lake, 1912, 1989; East Branch, 7267, 7268. New to the flora of the Northwest Territories. Arenaria Rossii R. Br. Ricuarpson Mrs.: West of Mackenzie Delta, 6659, 6792, 67934; Arctic Coast: Dolphin & Union Strait, Young Pt., C. A. E. 367B (as Sagina sp.) ; Bernard Harbor, C. A. E. 267 (as Alsine verna Bartl. var. rubella (Wahlenb.) Lange, Macoun & Holi in Rept. Can. Arct. Exp. 5A: 12. 1921). Arenaria sajanensis Willd. RicHarpson Mrts.: West of Mackenzie Delta, 6794, 7333 (western limit); Great BEar LAKE: Numerous collections; KEEwatin Distr.: Kazan R., 5794; Tha-anne R., 5575. 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 31 An eastern species, apparently common west to the Mackenzie. New to the Mackenzie District. Arenaria uliginosa Schleich. GrEAT BEAR LAKE: Smith Arm, 5077; Cape McDonnel, 5140; McTavish Arm, 3655; KEEWATIN Distr.: Kazan R., 5759A, 6629; Mistake Bay, 5661. New to the Mackenzie and Keewatin Districts. Cerastium alpinum L. Contrary to some writers who have stated that C. alpinum is circumpolar, in North America it appears to reach west only to the 120th meridian, where it is replaced by C. Beeringianum Cham. & Schlecht. Cerastium arvense L. MACKENZIE River DELTA: Campbell Lake, 1987; Stringer 14257, 62271 (Can); GREAT Bear LAKE: Dease Arm, 4803. New to the flora of the Northwest Territories. - Sagina intermedia Fenzl. Arctic Coast oF YUKON TERRITORY: 7099; MACKENZIE River DELTA: Kittigazuit, 2482; Arctic Coast: Cape Dalhousie, 2747. New to the flora of the Mackenzie District. Sagina nodosa (L.) Fenzl. GREAT BEAR LAKE: McTavish Arm, 5175; Krewatin Distr.: Baker Lake, 6119. New to the Mackenzie and Keewatin Districts. Silene repens Patr. in Pers. Syn. 1: 500. 1805. Silene purpurata Greene, Pitt. 2: 229. 1892. RicHarpson Mrts.: West of Mackenzie Delta, 6791; MackEeNziIE River Detta: East Branch, 7026; Richards Isl., 2085, 7068, 7270; Eskimo Lake Basin: 3028. Silene repens is rare or occasional in sandy places east to the Eskimo Lakes. It is one of the few truly western species that have crossed the Mackenzie River. New to the flora of the District. Melandrium (Wahlbergella) Much has been written in the past, by students of arctic plants, on the taxo- nomic status of this notoriously critical genus and its members, and it is not with- out a feeling of presumption that the writer embarks upon a treatment of the American arctic and boreal members of the group. While in the eastern Ameri- can Arctic the problems, if not simple, at least are limited to a few species, in the far Northwest matters become more complicated and the writer finds that he cannot conscientiously “‘lump” the numerous entities found there with the apparently still insufficiently understood “eastern” or circumpolar species. Although most American authors (except Rydberg) maintain Melandrium (Wahlbergella) in the somewhat heterogeneous or “roomy” genus Lychnis, the writer, without offering any new argument in favour of the admittedly slight distinguishing features, prefers to follow the European authors who, with few exceptions, have long maintained Silene, Melandrium, Viscaria, Lychnis, and Agrostemma as separate genera in the tribe Lychnideae. A. T. Tolmatchev (Trav. Mus. Bot. Acad. Sci. U. R. S. S. 24. 1932) has given what appears a most satisfactory treatment of the genus in Eurasia, and in Komarov’s Fl. U. R. S. S. (6: 714-723. 1936) he has treated the Section Gastrolychnis, giving what seem workable keys and descriptions of 13 species occurring in the Soviet Union. KY SARGENTIA [4 From the discussion in some recent papers (Polunin, |. c., 180-186; Th. S¢- rensen, Medd. om Grgnland 101°: 26-33. 1933; P. Gelting, ibid. 101°: 47. 1934), it would appear that even the long-known East American species of Melandrium are not well-understood. Due to the numerous edaphic or ecological variations so common in this genus, herbarium material especially is not always easy to interpret if mature seeds are lacking. The writer is very familiar with M. af- fine, M. triflorum, and M. apetalum, which, in West Greenland, at the Danish Arctic Station at Godhavn, he has grown from seed and observed for many years in cultivation as well as in the field. He has also seen and studied the genus in the field from Labrador to Alaska. From mountains of the west he has seen the material of Melandrium in the Gray Herbarium and in the National Her- barium of Canada, but, lacking field experience with the genus in the Rocky Mountains, he prefers to limit the following treatment to the arctic and subarctic plants. While it is realized that some ecological or edaphic forms, if without mature seeds, may not fit into the following key, it is offered, nevertheless, in the hope that it may be of service in all but extreme cases. An excellent key to M. affine, M. triforum, and M. apetaluim is given by Abromeit in Bibl. Bot. 8 [Heft 427]: 15. 1899. Key To MELANDRIUM IN ARCTIC NortH AMERICA A. Seeds large, 1.5-2.4 mm. across, with conspicuously inflated wing; calyx, at least in anthesis, conspicuously inflated. B. Seeds 1.8 mm. or over; calyx much inflated; petals barely exserted. C. Flowers solitary (very rarely two), nodding throughout anthesis and erect only when capsule is mature; calyx thin, papery, dark purple, puberulent, glutinous, almost globular; petals purple (wet places: circumpolar, arctic) ............. M. apetalum. C. Flowers 1 or 2, lateral, long-peduncled; calyx pubescent, urceolate; petals pale rose (alpine gravelly slopes, Norton Sd., Alaska) .................. M. macrospermum. B. Seeds less than 1.8 mm. across; petals much exserted, 114 to 2 times as long as the calyx. D. Perennial; stems erect, stiff, often purplish above, 15-25 cm. high; cauline leaves linear-obtuse, smaller than the basal leaves; calyx mostly dark purple, urceolate in fruit, 1-1.6 cm. long and 0.6-0.8 cm. in diam.; petals milky white (circumpolar, RPC): , st. wag ats ia ane arene ate area ath oho weeds SRS bb BEA lpi MCE ONE aaah 4B As M. affine. D. Short-lived perennial; stems green throughout, slender, more or less fractiflexed at the nodes, 30-50 cm. high; cauline leaves well-developed, attenuate; flowers few, long- peduncled; calyx green, campanulate, in fruit 0.6-0.8 cm. long, 0.6-0.7 cm. in diam. ; petals rose; seeds small, cuneate, with a narrow wing (riverbanks; Yukon— MACHONZIC FR.) 6 c1505sh Sing cctiants wrslaieie se Wala any adn d bw RE ee b4SRR M. Taylorae. A. Seeds small, 1 mm. or less in diam., wingless; calyx in anthesis scarcely or at least not conspicuously inflated. FE, Fruiting calyx cylindric, with green nerves; inflorescence many-flowered, the flowers on long, conspicuously appressed pedicels; plant tall, stiffly erect (plains: Sask.—Rocky Mts., north to Mack.) ....... 0... ccc ccc acc c cee cece a secesecceses es. M, Drummondii. i. Fruiting calyx urceolate; inflorescence few-flowered. I’. Seeds cordate or reniform in outline, strongly tuberculate; stems robust, erect- ascending, mostly dark purple above; flowers sessile or very short-peduncled (en- demic of Greenland) ........ 0... ce cece cece e cee ee ee ees eeeveeaeceeaee M. triflorum. F’, Seeds angular, more or less prismatic, merely granulate or rugose; stems erect, green throughout, rather weak; basal leaves soon withering; flowers pedunculate (Mac- kenzie Delta—Arctic Coast; Siberia) ........... 0c. ccc cece eee cece M. taimyrense. Fk, Fruiting calyx narrowly campanulate; seeds somewhat lustrous, dark brown, granulate at least on the dorsal side; stems ascending, 1- to several-flowered (dry limestone ledges; endemic of Great Bear Lake) .......000. 0.00 ccc ccc ee eee M. Ostenfeldii. 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 33 Melandrium apetalum (L.) Fenzl in Ledeb. Fl. Ross. 1: 326. 1842. Lychnis apetala L. Sp. Pl. 437. 1753; Fl. Dan. 5: tab, 806, 1781. Lychnis montana Wats. in Proc. Am. Acad. 12: 247. 1877. Lychnis attenuata Farr in Contr. Bot. Lab. Pa. 2: 419. 1904. Lychnis nesophila Holm in Rep. Sp. Nov. 3: 338. 1907. To the full discussion recently given by Polunin (1. c., 184-186), the writer has little to add beyond that M. apetalum, unlike the other species, is invariably found in moist places, such as wet tundra, moist meadows, or even in grassy places bordering brackish lagoons, in wet moss by alpine streams, or in wet clayey places on moraines almost to the snow line. The flowers are solitary, or very rarely there is a weakly developed second flower from the upper node, nodding in anthesis and erect only when the capsules open. The stems are weak and often somewhat flexuous. It is so well-marked that confusion with other species seems scarcely possible. Simmons (Phytogeogr. Arct. Am. Arch. 84, 1913) and later Polunin, 1. c., have pointed out that L. nesophila Holm is merely an extreme and abnormal variation. Lychinis montana Wats. and L. attenuata Farr seem very near M. apetalum (L.) Fenzl. The first is said to have the flowers erect or slightly nodding in anthesis; the latter seems scarcely distinguishable. Distribution: Common across arctic North America, south to the tree line, north to Ellesmere Island; high mountains of Alta., B. C., and Yukon; W. Greenland south to Disco Bay; E. Greenland south to Scoresby Sound. General Distribution: Circumpolar (not in Iceland), arctic-alpine. Melandrium macrospermum A. E. Porsild in Rhodora 41: 225, tab. 552, fig. 1-3. 1939. Known thus far only from the type locality: Alaska: Norton Sound, volcanic hills back of Unalaklet, 4. E. & R. T. Porsild 1147. Melandrium affine J. Vahl in Fl. Dan. 14: 5. 1843. Melandrium pauciflorum (Ledeb.) Ostenf. in Medd. om Gr¢gnl. 65: 173. 1923, nomen con- fusum. Lychnis affinis J. Vahl in Fries, Mant. 3: 36. 1842; Fl. Dan. 13: tab. 2173. 1836 (sub nom. L. triflora). Lychnis furcata Fern. in Rhodora 34: 22. 1932, non Viscago furcata Raf. Autikon Botanikon 28. 1840. Lychnis affinis Fries, Polunin, Bot. Can. E. Arct. 183. 1940. Lychnis ? Kingii Wats. in Proc. Am. Acad. 12: 247. 1877. Lychnis triflora Sommerf. in Mag. f. Naturv. 21: 151. 1824, non R. Br. Rafinesque, |. c., in 1840, described a plant from “Labrador and Hudson Bay” under Viscago furcata, which Fernald, |. c., takes to be Melandrium affine J. Vahl (1843), and for which he substitutes the combination Lychnis furcata (Raf.) Fern. to cover the plant of Greenland and N. America (but not that of Europe). Rafinesque’s plant, however, cannot have been a Lychnis (Melandrium), for Rafinesque, according to Fernald, 1. c., says: “Remarkably like the last plant [Physocarpon vespertinum Raf., based upon Lychnis vespertina Sibth.], but a real Silene not dioical and with 3 styles, smaller 4 to 6 inches high, calix and petals shorter, incarnate.” Rafinesque’s plant then is a true Silene and not a Melandrium, nor a Lychnis in the restricted sense (Gastrolychnis), both of which have five styles. Fernald’s second suggestion that Vahl’s plant from Greenland (and also that of N. America) differs from Fries’ plant from Scandinavia seems hardly tenable in the light of more detailed historical data. This history is briefly as follows (notes kindly supplied by Dr. Morten P. Porsild). 34 SARGENTIA [4 Lychnis triflora was mentioned by name (Ross’ Voy. Disc. App. p. 142. 1819), but not properly described, by Robert Brown, from poor material from N. W. Greenland. The Norwegian botanist, S. C. Sommerfeldt, in 1824, described a plant under that name collected at Godhavn, Greenland, by his friend Schwabe, Governor of North Greenland, taking it to be Robert Brown’s plant because its 3-flowered inflorescence suggested this and because it was clearly not one of the two species of “Lychnis” (Melandrium apetalum and Viscaria alpina) then known from Greenland. Melandrium (Lychnis) triflorum, however, does not grow at Godhayn proper, where Schwabe most likely made his collection, whereas M. affine is very common there. Sommerfeldt’s plant is undoubtedly M. affine, and the first unequivocal description of true MW. trifloriam was by J. Vahl in Fl. Dan. (1843). Jens Vahl, from 1828 to 1832, traveled extensively in S. W. Greenland, where he had been specifically instructed by Hornemann to secure materials for Flora Danica. But not until 1833, when Vahl came to North Greenland, did he see any Melandrium. At Godhavn he at once col- lected what he also thought was Brown’s “Lychnis triflora,” and a beautiful plate of this plant was in due course published under that name by Hornemann, FI. Dan. 13: tab. 2173. 1836. When Vahl visited the east coast of Disco Island and the Nugssuaq Peninsula, he realized that there was still another plant which at last was the real ‘‘Lychnis triflora’ of Brown. The Godhayn plant which had been illustrated in Flora Danica (tab. 2173) erroneously as Lychnis triflora, was by Vahl (in his unpublished manuscript and field notes preserved in Copen- hagen) then called “affine,” thereby suggesting a close “affinity.” Later, in 1843, J. Vahl himself published and edited Fl. Dan. (tab. 2356), showing the true Melandrium triflorum (R. Br.) J. Vahl, and correcting the earlier mistake in a footnote which reads: “Obs. L. triflora Fl. Dan. ¢. 2170 [obviously a misprint for 2173] est Melandrium affine J. Vahl msc. (seminibus—reticulatis marginatis, marginibus inflatis).”’ Lange (Nomenclator Fl. Dan. 137. 1887) later repeated this correction. Vahl did not then redescribe the plant, which, in his field notes in Greenland, he had called “affine,” because his friend Elias Fries (Mantissa 3: 36. 1842) had already done it for him the year before. From 1838 to 1839, J. Vahl joined Gaimard’s Expedition to Spitsbergen. He left the expedition in 1839 at the small town Bosekop in Alten Fiord (Finmark District of northern Norway), where he spent some time before returning to Copenhagen. Here, on the banks of Alten River, Vahl rediscovered his new species of Melandrium from Greenland, and specimens from his collection are in the Copenhagen herbarium. While at Bosekop, Vahl undoubtedly met the botanist Laestadius, who was a missionary or minister to the Laps, and who lived and worked not far from there. He already knew Vahl’s plant and had named it Lychnis Dorothea in honour of its original local collector, Vahl’s hos- tess at Bosekop, Madame Dorothea Klerck. Part of Laestadius’ type, which he gave to Vahl, is in Copenhagen and bears the following inscription on the label: “Lychnis Dorothea. Florens Alten Mad. Klerck legit. Obs. an vera Lychnis affinis tua est? Si habes parva specimina mittas. Tue Lychnis affinis habet calicem valde inflatam.”’ The plants cited by Elias Fries, 1. c., therefore, are those collected “ad Alten Finmarkiae” by Laestadius and J. Vahl, not by his father Martin Vahl as sug- gested by Fernald, |. c. Martin Vahl’s journey to Lapland took place in 1787-88, and he died on December 24, 1804, at Copenhagen (See O. Dahl, in Nyt Maga- 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 35 zin f. Naturvidensk. 59. 1921). Fries maintained a close correspondence with both Laestadius and J. Vahl and it was clearly his intention to describe J. Vahl’s plant. His citation of J. Vahl’s ‘Fl. Gr. Msc.” proves that Fries accepted the latter’s view that the plant of Greenland and Lapland belonged to the same species rather than that he (Fries) made a “misidentification of the Greenland plant of J. Vahl” (Fernald, 1. c.). The writer is unable to see any consistent differences between the plant of northern Scandinavia and that of Greenland and N. America. Except in the seed character, M. affine is subject to variations according to habitat. The dif- ferences given by Fernald, 1. c., in his comparative description of Lychnis furcata and L. affinis are not very large and all seem to be within the normal range of variation in this species. The seeds illustrated by Fernald (figs. 5 and 6) vary somewhat in shape, but in size, texture, and width of wing they are really quite alike. Figure 3, showing a flowering calyx from Torne Lapmark, is not typical M. affine but probably M. affine ssp. angustiflorum (Rupr.) Tolm. Fernald’s suggestion that the plant of N. Scandinavia is distinct from that of Greenland and N. America is further weakened by the fact that Polunin, 1. c., reports, under Lychnis affinis Fries, a plant collected by him on the west coast of Hudson Bay (Churchill), “inseparable from examples of this species gathered in northern Scandinavia, as Professor Fernald was kind enough to confirm for me.” The Churchill plant, of which the writer, in 1930, collected abundant material, and of which he has seen several other collections, approaches M. Tay- lorae (Robins.) Tolm. in its small, greenish-striped fruiting calyces and in the exceptionally well developed cauline leaves, but its seeds are indistinguishable from those of M. affine. Near it, on drier ground, grew typical M. affine. Melandrium affine is normally found in not too wet sandy or gravelly places near the seashore, on lake and river banks, and on rocky ledges. It does rather well in cultivation and responds generously to fertilizers. Distribution: Rare or occasional in northern Alaska and on mountains of the interior; common across arctic Canada, south to the tree-line, north to Ellesmere Isl.; in mountains of western United States (“Lychnis Kingii Wats.”) and Canada; common in west Greenland south to 65° 38’ N., and on the east coast south to 69° 30’ N. General Distribution: Circumpolar (not in Iceland), arctic-alpine. Melandrium Taylorae (Robins.) Tolmatchev in Trav. Mus. Bot. Acad. Sci. U. R. S. S. 24: 267. 1932. Lychnis Taylorae Robins. in Proc. Am. Acad. 28: 150. 1893. Lychnis affinis sensu Polunin, Bot. Can. E. Arct. 183. 1940, non Fries. To Robinson's short but very clear description it may be added that M. Tay- lorae, unlike the other species treated here, is a “short-lived perennial” which generally dies after the second flowering. The weak, fractiflexed or ‘‘zig-zag”’ stem, as well as the calyx, are viscid with short, glandular hairs; the basal leaves are lanceolate, short-petiolate, glabrate to almost glabrous except for the soft ciliation of the margins; the stem leaves are well-developed and attenuate. The lower flowers are borne on long peduncles and are nodding when young; the capsule is very short, less than 1 cm. long, and almost as wide. The seeds are small, triangular or wedge-shaped in outline, with a narrow but distinct wing, wrinkled but not granulate or tuberculate. Melandrium Taylorae, by its very short capsules, long fruiting peduncles, and the very characteristic seeds, is well-distinguished from M. affine. Polunin, 1. c., who had perhaps seen only the type (a poor specimen), or who had been 36 SARGENTIA 14 inisled by Ostenfeld, who annotated it “L. pauciflora Ledeb.” without giving his reasons, reduced it to synonymy under L. affints Fries. Mackenzie River Detta: Dry slopes, Peel River, July 15, 1892, Miss E. Taylor (TYPE, G), 2647 (isotype, Can) ; north end of Richards Isl., 2212; south end of Richards Isl., 2084; Arctic Coast: Tuktuayaktoq, 7418; MACKENzIE River: Arctic Red River, Dutilly 16 (G) ; Great Bear LAKE: Etacho Pt., 1500’ elev., 3492; Bear River, 3266; YuKoN TERRITORY: Klondike River, John Macoun 58402 (Can), distr. as Lychnis triflora var. Dawsoni Robins. ; Coal Creek Hill, /. Funston 81 (G, US); Bonanza Creek, A. Eastwood 195, distr. as L. triflora (G); ALASKA: Kowak R., N. Alaska, McLennan, in 1884 (G); Eagle Camp (centr. Alaska), E. Scamman 699 (G); International Boundary, 65° 30’ N., 141° W., Mertie 35 (US), General Distribution: Arctic Coast from N. W. Alaska to a short distance east of the Mackenzie Delta, south to Great Bear Lake, lower Mackenzie River, Central Yukon and Alaska. Melandrium Taylorae generally grows in sandy or loamy soil on river and lake shores, or sometimes in rather dry meadows or in open copses. Melandrium Drummondii (Wats.) comb. nov. Lychnis Drummondii Wats. in Bot. King Exp. 37. 1871. ? Silene Drummonditi Hook. Fl. Bor.-Am. 1: 89. 1830. Distribution: Prairies and plains from Manitoba to British Columbia and southward, north to Lake Athabaska. Melandrium triflorum (R. Br.) J. Vahl, Fl. Dan. 14: tab. 2356. 1843. Lychnis triflora R. Br. in Ross’ Voy. Disc. App. 142. 1819. Vahl’s brief but very lucid description and the beautiful plate in FI. Dan., |. ¢., show that this plant is abundantly distinct from 1. affine, and the difficulty ex- perienced by some investigators, e. g., Sgrenson, |. c., in E. Greenland may indi- cate that true M. triflorwm is rare there. As far as the writer is aware, all records from America have proved erroneous, and he agrees with Simmons (Vasc. Pl. Ellesmereland 126. 1906), notwithstanding Polunin, 1. c., that J, triflorum must still be considered an endemic of Greenland. Melandrium tri- florum is related to MW. Ostenfeldti, M. Drwmmondii, and M. taimyrense in the West. Distribution: W. Greenland south to 65° 40’; E. Greenland south to 70° 30’. Melandrium taimyrense A. Tolmatchev in Trav. Mus. Bot. Acad. Sci. U. R. S. S. 24: 264, fig. 7. 1932. Stems strict, simple or branching, erect, hoary-villous, 16-25 cm. high. Leaves linear, about 6 cm. long, somewhat obtuse, canescent-hirsute. Flowers mostly 2 to 4, rarely solitary, calyx dark-lined, covered with grandular white hairs. Corolla white (or pale rose?), slightly longer than the calyx, the petals emargi- nate. Seeds rugose, somewhat granulate, yellowish brown (the writer’s trans- lation from the original). _ The writer has seen an isotype, from Taimyr Peninsula, 74° 50’ N., 106° E., Aug. 13, 1928, A. Tolmatchev 765 (in Herb. M. P. Porsild). It 1s in fruit and is perfectly matched by fruiting specimens in the writer's collection. To Tol- matchev’s description it may be added that the flowering calyx is narrow, about half as wide as long, much smaller and narrower than in MW. affine or M. triflorum. The petals of the flowering plant, of which Tolmatchev apparently did not have material, are actually considerably exserted. The young plants are gray, hoary- canescent with a short, dense pubescence. The basal leaves tend to become glabrescent in age and often wither even before the capsules open. MACKENZIE River Detta: Limestone hills N. of Campbell Lake, 1973; Campbell Lake Seagull Cliff, dry gravelly limestone slopes, 1988; East Branch, dry riverbanks, 7025, Arctic Coast: 6 miles east of Kittigazuit, wet clay, 2475; Atkinson Pt., 2597; Cape Dal 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 37 housie, 2744; Great Bear Laxe: North shore of Dease Bay, 4691; KEEwATIN Dtstr.: West branch of Thelon R., 63° N., July 5, 1900, J. W. Tyrrell 23143 (Can). Distribution: Sandy and gravelly places, from the Mackenzie Delta east to Thelon R., south to the north shore of Great Bear Lake. General Distribution: Taimyr Pen., N. W. Siberia; Arctic, Mackenzie and Keewatin. Melandrium Ostenfeldii sp. nov. Herba caespitulosa e radice multicipite crassa ascendens, innovationibus nu- merosis sterilibus rosulatis ; culmi 20-30 cm. alti, tenues, plus minusve divaricati aut ascendentes, nodis saepe geniculati, foliorum paribus 1 vel 2, saepius diminu- torum, facie lineari-obtusa praediti, pubescentia brevi glandulosa in toto vestiti ; folia basalia lineari-spathulata aut sublanceolata dense breviter pubescentia, adulta glabrescentia, petiolis diu persistentibus; flores 1-8, plerumque 3 vel 4, erecti vel pedunculis subdivergentibus fulti; calyx campanulatus, 1 cm. brevior, glanduloso-viscosus, viridis cum lineis subobscuris purpurascentibus ; petala alba quam calyce dimidio majora; semina minima, 0.8 X 0.6 mm. aut minora, angu- laria, subnitida, castanea, conspicue granulosa. Great BEAR LAKE: Dease Arm, Narakay Isl., 66° 45’ N., 119° 30’ W., rocky ledges of diabase, Aug. 1, 1928, A. E. & R. T. Porsild 4839 (type); foot of Dease Arm, mouth of Dease R., 4802; north shore, mouth of Haldane R., 5036; McTavish Arm, east shore, 3777; McTavish Arm, N. E. Fjord, 3676; McTavish Arm, foot of Edna Travers Bay, 5218; Mc- Tavish Arm, north shore, 5233; south entrance to Conjuror Bay, 3623. Melandrium Ostenfeldii is known thus far only from pre-Cambrian rocks of the east shores of Great Bear Lake, where it is common on acid diabase and gran- ite cliffs. It differs from M. taimyrense by its narrowly campanulate fruiting calyx and by its well-developed radical leaves that persist through the fruiting stage. It is no doubt related to M. triflorum, but, like that in IW. taimyrense, the flowering calyx is much narrower and the seeds are smaller and granulate rather than tuberculate. Stellaria calycantha (Ledeb.) Bong. MACKENZIE River Detta: East Branch, 6541, 7266; Esktmo Lake Basin: Setidgi Lake, 3144, 3178; Great Bear Lake: Keith Arm, 3465; Etacho Pt., 3489; McTavish Arm, 3894; north shore, 5232. Somewhat rare, due to scarcity of suitable habitats, north to the limit of trees or slightly beyond. Stellaria crassifolia Ehrh. Represented by many numbers in the writer’s collection, from Herschel Island to the west shore of Hudson Bay. Near its northern limit on the Arctic Coast, S. crassifolia, perhaps normally, does not produce fruit but is propagated by winter buds. New to the flora of the Mackenzie District. NYMPHAEACEAE Nuphar variegatum Engelm. Eskimo Lake Basin: Setidgi Lake, 3704, fruiting abundantly; McTavish Bay, east shore, 3704. Previously known in our region north to Great Bear Lake, where it is fairly common. RANUNCULACEAE Caltha natans Pall. KeewatTin Distr.: Lake on Tha-anne R., 60° 58’ N., 97° W., 5502. New to the flora of the Northwest Territories. 38 SARGENTIA |4 Anemone narcissiflora L. RicHarpson Mrts.: West of Mackenzie Delta, 6588, 6665, 6807, 7334. Common in alpine meadows above the timber line. New to the flora of the Northwest Territories. Ranunculus acris L. var. frigidus Regel. Ranunculus occidentalis Nutt. var. robustus Gray, Ostenf. in Gjéa Exp. 46. 1909; Macoun & Holm in Rept. Can. Arct. Exp. 5A: 13. 1921. RicHarpson Mrs.: 6591, 6810, above the timber line. Common along the Arctic Coast and in mountains west of the Mackenzie Delta. Our plant is 20-30 cm. high and perfectly glabrous except just below the flower. The flowers are very large, 2 to 3 cm. in diameter; the petals are obovate and very shiny. The achenes are large with a prominently hooked, persistent style. Ranunculus circinatus Sibth. ? Ranunculus subrigidus Drew in Rhodora 38: 39. 1936. MACKENZIE River De_ta: East Branch, 68° 40’ N., in 2-3 feet of water, 7248. The writer is unable to find any characters wherein the American plant differs from that of the Old World. The flowers, which are larger than in R. tricho- phyllus Chaix, are emerged during anthesis. Also it is later-flowering than that species. In the Mackenzie Delta it does not flower or mature fruit except in favorable seasons. New to the flora of the Northwest Territories. Ranunculus gelidus Karelin & Kirilow in Bull. Soc. Bot. Mosc. 15: 133. 1842. For complete nomenclature of this rare plant see Ostenf. in Gjoa Exp. 4. 1909. To the synonyms given there should be added R. Grayi Britton, in Bull. Torr. Bot. Cl. 18: 265. 1891, based upon R. pedatifidus Hook. Fl. Bor.-Am. 1: 18, tab. 8B. 1829, non Sm. The Gjoa Expedition collection came from King Point on the Arctic Coast of Yukon Territory. Abundant material was collected on the Richardson Mts. : West of Mackenzie Delta, 68° N., 136° W., on alpine peak, 4000’ elev., 6660. Ostenfeld, |. c., has shown that the plant so well-described and illustrated by Hooker, I. c., from specimens collected by Drummond in the Canadian Rockies, cannot be kept apart from Karelin & Kirilow’s plant from the Alatau Mountains of Central Asia. In his recent treatment, Benson (in Bull. Torr. Bot. Cl. 68: 657. 1941) apparently has overlooked Ostenfeld’s discussion, and he maintains the plant under FR. Grayi, stating that it is related to R. Eschscholtzii. The writer again is inclined to agree with Ostenfeld, who states that it belongs “to the auriconus-group, as also R. affinis and R. pedatifidus.” Our specimens grew in moist gravel with Douglasia arctica. On July 10th the achenes were mature and but a single flower remained. The petals are small, 4-5 mm. long, pale yellow, and barely exceeding the sepals. The flowering stems are 6-10 cm. high, somewhat arcuate; the peduncles are villous. The leaves are glabrous and glaucous, the old leaf-bases persistent and strongly fibrous, as are the strongly developed roots. The achenes are glabrous, light green, 2 to 2.4 mm. in diam., orbicular, flattened; the style is lateral, short-hooked, not at all as figured by Hooker’s Fig. 1, drawn from immature material. Ranunculus Pallasii Schlecht. Arctic Coast oF YuKoN TERRITORY: Shingle Pt., 6905; MACKENZIE River Detta: North end of Richards Isl., 2278; Arctic Coast: Tuktuayaktoq, 7419; Atkinson Pt., 2612; Liver- pool Bay, 3104; Eskimo LAKE Bastin: 2985, 3104. 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 39 Common in wet brackish marshes along the Arctic Coast east to Anderson River, often together with Hippuris vulgaris. The very showy, pure white or pale rose flowers are very fragrant. The plants blossom throughout the season, and flowers and mature fruits may be seen together. It seems very strange that such a keen observer as Richardson failed to observe this very conspicuous plant. It is unrecorded from Canada west of Hudson Bay, but it is common in arctic parts of Alaska. Ranunculus trichophyllus Chaix var. eradicatus (Laestad.) Drew. MacKENZIE River DELTA: Very common; Kittigazuit, 2489; Eskimo Laxe BAsIN: 3179; Great Bear LaKe: North shore, 5038; Dease Arm, 4695; McTavish Arm, 5270; KEEWATIN Distr.: Kazan R., 6036. The var. eradicatus seems always to prefer deeper water than var. typicus. In the Mackenzie Delta and at Great Bear Lake it was often observed to form a dense growth, flowering and fruiting abundantly, on the bottoms of lakes 10 to 12 feet deep. The flowers barely open and may be cleistogamous. In rapids of the Kazan River, in water 4 to 8 feet deep, it was likewise observed to flower and fruit. Ranunculus trichophyllus Chaix var. hispidulus Drew. Great Bear Lake: McTavish Arm, north shore, in a shallow, marshy slough, 5299. Ranunculus trichophyllus Chaix var. typicus Drew. MAcCKENzIeE River Detta: East Branch, 6919; Eskimo Lake Bastin: North shore of 2nd lake, 2983, 2984. ; Much less common than var. eradicatus and perhaps restricted to shallow, calcareous lakes. , PAPAVERACEAE Papaver alaskanum Hultén. Rare or occasional along the Arctic Coast east to Cape Bathurst. In most of the collections the flowers are sterile, i.e., the anthers are well-developed but the ovary is abortive. Where fertile flowers occur the immature capsules are frequently eaten by ground squirrels (Citellus parryt). Papaver microcarpum DC. RicHarpson Mrs.: West of Mackenzie Delta, alpine tundra, above timber line, 6669, 6813-6815 (the last is f. albiflora) ; Arctic Coast oF YUKON Territory: King Pt., 7179; Shingle Pt., Can. Arct. Exp. 180 (distr. as P. nudicaule L.); Mackenzie River: Ft. Nor- man, G. Hume 103443 (Can). Papaver microcarpum is apparently common in the mountains of Yukon Terri- tory but does not reach east of the Mackenzie. Papaver radicatum Rottb. GreAT Bear LAKE: Cape McDonnel, 5143; McTavish Arm, north shore, 5235; east shore, 3781. Papaver radicatum is the common species of eastern Arctic America and the Arctic Archipelago. It reaches west to Great Bear Lake and Cape Bathurst. West of Great Bear Lake it is replaced by the closely related P. alaskanum, which may perhaps be considered a western race. (See Porsild in Rhodora 41: 230-— 231, 1939). CRUCIFERAE Subularia aquatica L. Great Bear Lake: McTavish Arm, 3696; KerewatiIn Distr.: Lake on Tha-anne R., 60° 58’ N., 97° W., 6176. New to the flora of the Northwest Territories. 40 SARGENTIA |4 Thlaspi arcticum sp. nov. Thlaspi montanum sensu Hook. Fl. Bor.-Am, 1: 58. 1830, non L. Thlaspi alpestre L. var. purpurascens Ostenf. in Gj6a Exp. 47, tab. 3. fig. 17. 1910, quoad spec., non 7. purpurascens Rydb. in Bull. Torr. Bot. Cl. 28: 281. 1901. Herba perennis, radice in caudice multicipite ramoso terminata; folia basalia subglauca, carnea, glabra, spathulata, integra, in petiolum laminae aequilongum sensim fastigiata, 1-2.5 cm. longa, 0.5-0.8 cm. lata, nervo medio prominente, nervis lateralibus inconspicuis; culmi floriferi singuli aut pauci, glabri, in anthesi 3-5 cm. alti, maturi valde elongati, 10-18 cm. alti; folia caulina 3-5, linearia, - sessilia, basi truncata ; inflorescentia capitata 2—25-flora, fructifera valde elongata, pedicellis valde divaricatis, longitudine siliquas aequantibus; petala alba, circa 4.5 mm. longa, unguiculata, nervosa, sepalis viridescentibus duplo longiora; fila- menta brevia crassaque, antherae breves ellipsoideae; siliquae maturae 6-7 mm. longae, 2-2.5 mm. latae, subcurvatae, cuneato-clavatae, exalatae, obtusae vel apice subacutae, haud emarginatae, valvis valde carinatis; stylus filiformis, circa 1 mm. longus; septum incompletum scissura magna longitudinali instructum ; seminibus 1.5—2 mm. longis, 0.8-1 mm. latis, levibus, brunneis. Arctic coast and foothills west of Mackenzie Delta, west to Herschel Island. Arctic Coast oF YUKON TERRITORY: Kay Pt., 69° 12’ N., 138° 30’ W., very rare on dry tundra ridges, July 23-25, 1934, A. FE. Porsild 7144 (type); King Pt. (about 138° 30’ W.), June 20-28, 1906, Gjéa Exp. (Copenhagen, Can); Herschel Isl., 69° 35’ N., July 18, 1906, Gjéa Exp. (Copenhagen); same place, 1893, /. O. Stringer 14257 (Can), distributed as Arabis lyrata L. var. ambigua. Our specimens are in fruit, while those of the Gjda Expe- dition are in anthesis. Ostenfeld, |. c., suggests that the Gj6a Exp. plant is 7. purpurascens Rydb., from high mountains of Colorado and Arizona. He reduces it to a variety of the European T. alpestre, which, however, as already pointed out by Rydberg, l. c., 1s not indigenous to North America. Furthermore, 7. alpestre is not perennial; Thellung (in Hegi, Ill. Flora v. Mitteleuropa 4': 122) says of it: “Plant mostly bi- or triennial and, having once flowered it dies” (translation of Thellung’s German text). Thlaspi arcticum, as well as all native Thlaspi of the Rocky Mountains, is perennial. In his revision of the genus, Payson (Univ. Wyom. Publ. Bot. 1: 6. 1906) cites Hooker’s Fl. Bor.-Am. but not Ostenfeld, 1. c., and makes no reference to the arctic plant, nor does he appear to have seen specimens of it. Thlaspi arcticum differs from T. purpurascens and other related species by its much shorter style and few-leaved flowering stems; it is perhaps most closely related to T. cochleariforme DC., of mountains of western Siberia. One sheet in the Nat. Herb. of Canada: “Montes uralenses polares (lat. bor. 65° et 66° 30’), July 17, 1926, B. Gorodkov,” distr. by Herb. Inst. Bot. Acad. Sei. U. R. S.S., differs from T. arcticum by the marcescent old leaf-bases, pu- berulent flowering stems, and the barely elongated fruiting raceme. Rorippa obtusa (Nutt.) Britt. MACKENZIE River DeELta: East Branch, riverbanks and meadows, 7225. New to the flora of the Northwest Territories. Thellungiella salsuginea (Pall.) O. E. Schulz. Sisymbrium salsugineum Pall. in Ledeb,. Fl. Alt. 3: 145. 1831. ? Turritis diffusa Hook. Fl. Bor.-Am. 1: 41. 1829. A single collection of what appears to be this rare species was made on the Arctic Coast, east of the Mackenzie, Liverpool Bay, 2897. Our plant is appar- ently annual; it has clasping stem-leaves and entire radical leaves. 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 41 Cardamine digitata Richards. in Franklin’s Journ. App. 743. 1823; Hook. Fl. Bor.-Am. 1: 45. 1829, excl. pl. Bering Str.; Macoun & Holm in Rept. Can. Arct. Exp. 5A: 14. 1921; Holm, ibid. 5B: 37. 1922. Cardamine hyperborea Schulz in Bot. Jahrb. 32: 550. 1903, pro min. parte; Ostenf. in Gjéa Exp. 48. 1909; Simmons, Phytogeogr. Arct. Am. Arch. 91. 1913; Polunin, Bot. Can. E. Arct. 231. 1940. See Porsild in Trans. Roy. Soc. Can. 3rd. ser. sect. V, 32: 31-32. 1938. When Schulz, |. c., united the genus Dentaria Tourn, with Cardamine Tourn., because of Dentaria digitata Lam., it became necessary to rename Cardamine digitata Richards. But, as pointed out by Holm, |. c., there is very good au- thority for maintaining the genus Dentaria as distinct from Cardamine, and very few authors, indeed, have followed Schulz. Furthermore, as pointed out by the writer, 1]. c., Schulz appears to have misunderstood C. digitata Richards., and the plant which he discussed under C. lryperborea appears largely to have been C. Blaisdellii Eastw. of the Bering Sea region. Certainly the plant he de- scribed is not Richardson’s plant at all. Cardamine digitata Richards. still remains a very rare plant in most herbaria, and, as far as the writer is aware, it has never been illustrated. It is a plant of not-too-wet tundra, reaching south to the tree line, but apparently not truly arctic, since at Cape Dalhousie, in latitude 70° N., specimens were in beginning anthesis in the early half of August (2766). Cardamine digitata prefers rather dry, peaty soil and often grows on the sides of large “niggerheads.” The slender rhizomes and the vegetative reproduction have been well described by Holm, 1. c.; the rhizomes are deeply buried in the turf and can be dug out only with the greatest care. Fructification is poor, and as a rule but half of the ovaries are fertile. Mature siliques are 30 to 40 mm. long and 2 mm. wide, gradually tapering into a slender style 2 to 3 mm. long and containing from 5 to 8 seeds. The seeds are oval, somewhat flattened, 1.4 mm. long and 0.8 to 1 mm. wide, greyish-brown, smooth. Richardson, |. c., is mistaken when he says of the color of the flowers: “alba vel purpureo tincta.” The petals invariably are milky-white. Ostenfeld, 1. c., has pointed out also that Richardson’s description of the leaves, “‘folia digitatim pinnata,” is incorrect and that the leaves actually are pinnate. Cardamine digitata appears to be an endemic of North America, rare in north- western Alaska, fairly common along the arctic coast of the Yukon and Mac- kenzie Districts east to Hudson Bay, north to Banks and Victoria Islands, and south to the tree line. AtAska: Teller, Walpole 1465, 1561, 1810, 1890 (US); Mendelhall, Dall R. trail, Dall Cy. (US); Arctic Coast or YuxKon Territory: 7182; Herschel Isl., Can. Arct. Exp. 248; NorTHWESTERN Yukon: 66° 02’ N., 141° W., D. D. Cairnes 83045 (Can); RicHaRDsON Mrs.: 6678; Mackenzie Detta and Arctic Coast: numerous collections; GREAT BEAR Lake: Dease Arm, 4736; Cape McDonnel, 5146; Keewatrn Distr.: Kazan R., 5812; Baker Lake, 6123; Mistake Bay, 5672. Draba ? aurea M. Vahl. Mackenzie River Detta: Campbell Lake, dry hillsides on limestone, 1995; East Branch, sandy river banks, 6996. These two plants, except for the entirely white petals, are a perfect match for ' D. aureiformis Rydberg, considered by Fernald (in Rhodora 36: 300. 1934) in- separable from D. aurea M. Vahl. By Mrs. E. Ekman both (in herb.) were named D. Thomasii Koch (D. lanceolata Royle), which most certainly they are not. The writer is very familiar with D. aurea in Greenland and in the East, 42 SARGENTIA |4 and it may be significant that in life this species did not suggest itself to him. Our specimens are biennial or at least appear to be hapaxanthic. The flowering stems are branching and the flowers white, the young siliques stellate-pubescent and the styles 1 mm. long and slender. Draba crassifolia Grah. MACKENZIE River Detta: North end of Richards Isl., 2238; Keewatin Dtstr.: Kazan River, 63° N., 6037. New to the flora of the Northwest Territories. Draba incerta Payson. Draba ? glacialis B Hook. Fl. Bor.-Am. 1: 51. 1830. GREAT Bear Lake: J. M. Bell 22874 (Can). In the Gray Herbarium is a Franklin Expedition plant labelled: “C¢. H*. [Cumberland House] to Bear L.—R[ichardson]. June-—Aug. Draba glacialis 8.” It is in anthesis, without fruit, but seems typical D. incerta. New to the flora of the Northwest Territories. Draba luteola Greene. Great Bear Lake: Keith Arm, 3460, 3462; north shore, 4739, 4946, 5004; McTavish Arm, 5272; Leith Pt., south shore, 3580; Coronation GuLF: Can. Arct. Exp. 426 (un- named) ; Britis Corumsra: Mt. Selwyn, 56° 01’ N., H. M. Raup & E. C. Abbe 3915, 4104 (distr. as D. borealis DC.). The material listed above is a good match for D. luteola Greene, otherwise known from Colorado and Utah. It is the plant which the late Mrs. E. Ekman (Sv. Bot. Tidskr. 28: 80. 1934) called a hybrid: D. aurea & daurica and “prob- ably also Dr. arabisans Michx.” (Fernald, 1. c., pb 25%, Our plant grew on sandy, calcareous beaches. In life the petals were white or creamy white, not yellow. In most of our material the lower part of the stem and basal leaves are tinged with purple. Draba oligosperma Hook. Fl. Bor.-Am. 1: 51, 1829. Great Bear Lake: Etacho Pt., 1500’ elev., 3498: McTavish Arm, east shore, greenstone hills, 900’ elev., 5302. Characteristic for this species is the very remarkable elongation of the inflores- cence, which, as pointed out by Hooker, |. c., takes place as early as anthesis. So far as the writer is aware, this is not paralleled in any other of the dwarf arctic or boreal Drabas. Schulz incorrectly places D. oligosperma in the section “Chrysodraba,” describing the petals as being yellow (“flava”); Hooker cor- rectly states that they are white (“petala alba”), but in drying they may turn a pale yellow. Part of Hooker’s type, “Summit of limestone hill, Mack. R. 68° Dr, Rich.,” is in the Gray Herbarium. It is a somewhat more slender plant than the writer’s. Draba Palanderiana Kjellm. in Vega Exp. Vet. lakt. 2: 45. 1883. Ricnarpson Mrs.: West of Mackenzie Delta, gravelly slopes, 4000’ elev., 6595, 6071, 6673, 6675. This curious Draba, which is a close match for Kjellman’s plant from Seward Pen., Alaska, was interpreted by the late Mrs. Ekman as D. nivalis xX pilosa. It forms large, loose rosettes. The inflorescence is capitate in anthesis; the flowers large, white, but turning yellow in drying. The siliques are lanceolate, glabrous, with a long style. A specimen from the type locality, Teller [Port Clarence], Alaska, Walpole 1402 (US), is 20 cm. in diam. with several hundred flowering scapes. 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 43 Draba praealta Greene, Pitt. 3: 306. 1898. Great BEAR Lake: Smith Arm, 5003, 5005, 5006. New to the flora of the Northwest Territories. Smelowskia calycina (Steph.) C. A. Mey. RicHarpson Mrs.: West of Mackenzie Delta, barren alpine ridges, between 2000 and 3000’ elev., 6596, 6680, 6821. Although the plants flower profusely, the fruiting is very poor, and but a single silique, containing one good seed, could be found. Not previously recorded from the Northwest Territories. Arabis Drummondi Gray. RicHarpson Mrts.: West of Mackenzie Delta, 6819; Great Bear Lake: Etacho Pt. 1500’ elev., 3494. New to the flora of the Northwest Territories. Arabis Holboellii Hornem. See Rollins in Rhodora 43: 440. 1941. GREAT BEAR LAKE: Bear River, 3364; Smith Arm, 5080; McTavish Arm, east shore, 3705, north shore, 5304. Our specimens are perennial and a perfect match for the plant in its type lo- cality (Godhavn, Greenland), and they represent var. typica Rollins, recorded by him, |. c., in the far northwest, from Yukon and Alaska. New to the flora of the Northwest Territories. Arabis Hookeri Lange. MACKENZIE River Detta: Campbell Lake, 1920, 1921, 1994; East Branch, 6892, 7395; Richards Isl., 2089; Kittigazuit, 2362; Arctic Coast: Atkinson Pt., 2623; Great Bear LAKE: McTavish Arm, north shore, 5202; east shore, 5305A. Our specimens are very much like the plant of Greenland and are all biennial. Arabis Hookeri flowers very early and our nos. 1920, 1921, and 1994, collected June 15-17, 1927, have almost mature siliques. As in Greenland, A. Hookeri is a pronounced calciphile and a dung-lover. In the Mackenzie District it is often found below bird cliffs and around the burrows of ground squirrels. The seeds are strongly mucilaginous when wet. The cotyledons are incumbent. Arabis Hooker is a rare plant in most American herbaria, and, since originally collected by Richardson on the “Shores of the Arctic Sea, between long. 107° and 130°,” it has been collected and recorded but a few times from North America (Yukon) outside of Greenland. There can be little doubt, however, that another plant, Halimolobos virgatus (variously known under Sisymbrium, Stenophragma, Pilosella, Arabidopsis), is identical with it. The arctic plant also has been transferred to Halimolobos as H. Hookeri (Lange) Schulz. Rol- lins, 1. c., who points out that the reason for this is chiefly the incumbent coty- ledons, brings it back to the original specific name as Halimolobos mollis (Hook.) Rollins. While the transfer to Halimolobos may be justified, it is, nevertheless, interesting to note that another recent monographer of the genus points out that in Arabis the “cotyledons vary from accumbent to incumbent or they may be quite oblique” or even “accumbent and incumbent on the same plant” (Hopkins in Rhodora 39: 72. 1937). Be this as it may, Halimolobos virgatus, with its host of “aliases,” should be added to the synonyms of A. Hookeri. Arabis lyrata L., Raup in Jour. Arn. Arb. 17: 259, 1936. GREAT BEAR LAKE: McTavish Arm, 3607, 3646, 3657, 3732, 5273. Raup, |. c., suggests that Sisymbrium arabidoides Hook., Fl. Bor.-Am. 1: 63. 1830, “to Lat. 68°,” is our plant. But Hooker, 1. c., stresses the fact that his 44 . SARGENTIA [4 plant is annual and has incumbent cotyledons. Our plant is glabrous through- out, is distinctly perennial, and has accumbent cotyledons. Braya Ostenfeld very properly stated that “there is much confusion concerning the arctic species of Braya” (Gjoa Exp. 17. 1909). One reason why this is so is that even in the Arctic these plants are rare and of a most peculiar and spotty distribution, suggesting special soil requirements. They are therefore not at all well represented in most herbaria. Four species were collected by members of the Franklin Expeditions and three were first described from our region. Braya purpurascens (R. Br.) Bunge is perhaps the best-known and most widely distributed. Study of this species in cultivation is probably necessary in order to understand its relation to related species such as B. Longii Fern., B. americana (Hook.) Fern., and the very near B. Thorild-Wulffii Ostenf., as well as B. purpurascens var. dubia. Braya glabella Richards. has been the cause of most of the misunderstanding and it has been interpreted in different ways by almost every writer who has dealt with it. Gelert (Bot. Tidsskr. 21: 292. 1898), who examined the type in the British Museum, was the first to point out that, contrary to the opinion of most previous writers, B. glabella was quite distinct from B. purpurascens. He claimed also that it was inseparable from B. alpina Sternb. & Hoppe of Central Europe, Greenland, and northern Scandinavia. Rouy (Ill. Pl. Eur. Rar. fase. 11. 84, tab. 254. 1899), who apparently had not seen Richardson's type of B. glabella, considered it more closely related to B. purpurascens and quite distinct from the plant of Greenland and Scandinavia, for which he proposed the new name B. linearis Rouy. Ostenfeld, |. ¢., agreed with Rouy that the plant of Greenland and Scandinavia was distinct from B. alpina of Central Europe but thought that Gelert was correct when stating that the plant of Greenland and Scandinavia was inseparable from that of arctic Amer- ica. Ostenfeld, therefore, correctly concluded that Rouy’s name was _ super- fluous and that Richardson’s B. glabella must stand. Alm (in Acta FI. Suec. 1: 250-252. 1921), who again revised the group, came to the same conclusion. Schulz (in Pflanzenr. 86 (IV. 105) : 230. 1924), on the other hand, agreed with Rouy, saying about B. glabella: “haec species a B. alpina et B. lineari habitu robustiore et siliquis crassioribus distinguitur,’ thereby showing that he, after all, was perhaps not referring to Richardson’s plant but to B. pilosa Hook., al- though he later reduced the latter to a “forma” of B. purpurascens. Braya pilosa Hook. has been largely ignored and the writer was particularly pleased to secure abundant material in the fruiting stage from the type locality of what he considers must be B. pilosa. Braya glabella Richards. in Franklin’s Journ. App. 743. 1823; Hook. Fl. Bor.-Am. 1: 65. 1830; Ostenf. in Gjoa Exp. 17. 1909; Alm in Acta Fl. Suec. 1: 247-262, tab. 16, figs. 4-7. 1921. Braya alpina Gelert, Bot. Tidsskr. 21: 291, fig. 2. 1898, not Sternb. & Hoppe. Braya linearis Rouy, Il. Pl. Eur. Rar. fase. 11. 84, tab. 254. 1899. To Richardson’s description it may be added that the whole plant is often tinged with purple. The stem, pedicels, and petioles (the blades only when young) are sparsely clothed with short, white, adpressed simple or bifurcated trichomes. In anthesis the inflorescence is capitate and barely exceeds the leaves; later a very considerable elongation takes place and fruiting specimens 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 45 are 15 to 20 cm. high. The mature siliques are 1 to 1.2 cm. long, on very short pedicels, almost terete, about 1.5 mm. in diameter. The style is short and conical in distinct continuation of the silique. The valves are glabrous and tinged with purple. Each cell contains from 10 to 14 dark brown seeds in two distinct rows. The seeds are oblong, pointed, about 1 mm. long, finely pebbled. The root is a weak tap-root, lasting only a few years. Braya glabella flowers very early and specimens with mature fruits were collected on June 23-26, 1928. We saw this species but once, not far from the type locality (Copper Mts., north of Dismal Lakes). Great Bear LaKxe: North shore of Dease Arm, 4698 (fruiting specimens), 4699 (flower- ing specimens from edge of snow bank). General Distribution: Great Bear Lake, W. & E. Greenland and N. Scandinavia. Braya humilis (C. A. Mey.) Robins. Syn. Fl. N. Am. 1: 141. 1895. Sisymbrium humile C. A. Mey. in Ledeb. FI. Alt. 3: 137. 1831, but first published in Hook. Fl. Bor.-Am. 1: 62. 1830, credited to ‘““Ledebour, MSS. ined.” Sisymbrium arabidoides Hook. F1. Bor.-Am, 1: 63, tab. 21. 1830, at least in part. Pilosella Richardsonii Rydb. in Torreya 7: 159. 1907. Braya Richardsonii (Rydb.) Fern. in Rhodora 20: 203. 1918. Braya alpina sensu Macoun & Holm, Rept. Can. Arct. Exp. 5A: 13. 1921, non Sternb. & Hoppe. Braya glabella sensu Holm, Rept. Can. Arct. Exp. 5B: 35, 1922, non Richards. Meyer, |. c., states that the plant is biennial, while Hooker, |. c., says it is perennial. The latter view is correct with regard to our plant, but the root 1s ‘weak and the plant usually dies aiter the second flowering. Braya humilis is a variable species. On limestone cliffs along the Mackenzie there grows a gray-canescent, lilac-flowered plant which is Rydberg’s Pilosella Richardsonii (in his Rocky Mt. Flora, p. 341, placed under Arabidopsis with the synonym Sisymbrium humile Hook., non Meyer). On the Arctic Coast the plant looks somewhat different and has slightly smaller and pure white flowers. This I suspect is at least in part what Hooker, |. c., called Sisymbrium arabidoides and which he illustrated in tab. 27. , MACKENZIE River DELTA: Campbell Lake, 1991; Arctic Coast: North end of Richards Isl., 2228, 2234, 2235; Atkinson Pt., 2622; Cape Dalhousie, 2761; Liverpool Bay, 2890; Victorra Ist.: Can. Arct. Exp. 411 (distr. as B. alpina) ; GREAT BEAR LAKE: North shore, Haldane R., 5039; McTavish Arm, north shore, 5177; Bear River, Mt. Charles, 3294, 3331; MACKENZIE River: Bear Rock, near Ft. Norman, 1300’ elev., 3383. General Distribution: Circumpolar, arctic-boreal. Alaska, Yukon, Mackenzie south to Alberta, James Bay and southern Hudson Bay, Gulf of St. Lawrence, Newfoundland, E. & W. Greenland. Not in Iceland. Braya pilosa Hook. Fl. Bor.-Am. 1: 65, tab. 17A. 1830. Braya purpurascens f. pilosa Schulz in Pflanzenreich 86(1V. 105): 235. 1924. There are in the National Herbarium of Canada two sheets of what is undoubt- edly part of the type collection of Braya pilosa Hook. Both have labels written by J. M. Macoun. The first, 2779 (Can), reads: “Braya pilosa Hook. Sea shores, mouth of Mackenzie River, Lat. 70°, N. W. T. Richardson.” It con- sists of a branching caudex bearing a flowering scape from each fork. The flowers are very large, with petals 7 mm. long. The other sheet, 2778 (Can), is labelled: “Braya alpina Sternb. & Hoppe var. glabella Wat. Sea shores be- tween Mackenzie and Coppermine Rivers, N. W. T. Dr. Richardson.” A second label is marked: “Ex. Herb. Musei Brittanica. Mouth of River Mac- kenzie Arctic North Coast. Dr. Richardson.” The second plant shows a strong 46 SARGENTTIA {4 tap-root branching into 11 caudices each bearing a flowering scape from 6 to 8 cm. high. The flowers are slightly smaller than in 2179 and the whole plant is more hirsute. It is a perfect match for Hooker’s plate 17A and for his descrip- tion of B. pilosa. Fruiting specimens of what is undoubtedly B. pilosa were collected on the Arctic Coast east of the Mackenzie Delta, in the latitude of the type locality (latitude 70°) : North end of Richards Isl., 2229; Atkinson Pt., 2621; Cape Dal- housie, 2760; Liverpool Bay, 2889. Our plant was at first taken to be B. glabella, but when later this plant was collected near its type locality the difference at once became apparent. The leaves, scapes, pedicels, and the valves of the mature silique in our plant are sparsely clothed with a short, soft pubescence of simple and bifurcated gray hairs. The whole plant is green and but very rarely tinged with purple. The inflores- cence is elongated in fruit but less so than in B. glabella. Also the mature siliques are broader than in that species, distinctly broader below the middle and decidedly flattened, with a filiform style 1.2 to 1.5 mm. long. The seeds are similar to those of B. glabella but decidedly paler, 10 to 14 to each cell. Richard- son observed that the flowers are very fragrant, comparing the “smell of the blossoms to that of Lilac.” Braya pilosa apparently flowers very early, and, although the writer diligently searched for them during several seasons spent near the Mackenzie Delta, he never succeeded in finding flowering specimens, the reason being that the Arctic Coast, where the plant grows, is almost inaccessible during the early part of the season due to ice conditions. Franklin’s party reached the sea-coast on July 8, 1826, which must have been in an early season, for Richardson was able to pro- ceed east along the coast unhindered by ice. In 1928 the late Dr. M. O. Malte collected in the Hudson Strait region a Braya which he tentatively named B. glabella Richards. The Hudson Bay plant is definitely not B. glabella Richards. but is a close match for our material of B. pilosa. As in the west, it flowers very early and only fruiting specimens have been collected. Of the Hudson Bay plant the writer has seen the following collections: Northern Quebec: Wakeham Bay, M. O. Malte 120191 (Can, G) ; Southampton Isl.: WM. O. Malte 120677 (Can, G). On Baffin Island Malte collected another Braya which he called B. purpuras- cens Bunge (Lake Harbour, M. O. Malte 120294, Can, G). This differs from the Wakeham Bay and Southampton Island plants by the less elongated fruiting head, shorter and stouter styles, and by the shorter and more plump siliques. Polunin, |. c., also collected the latter plant at Lake Harbour (1721, Can), label- ling it a variety of B. purpurascens (R. Br.) Bunge. The writer has seen specimens of B. purpurascens from Greenland that are a close match for the Baffin Island plant (W. Greenland, Ingnerit Fj., 71° 7’ N., M.P.& R. T. Porsild, Can). Braya pilosa Hook. no doubt is more closely related to B. purpurascens than to B. glabella, but, because of its early seasonal appearance and very large and fragrant flowers, much elongated fruiting racemes, long and slender styles, and erect habit of growth, it is considered abundantly distinct. Braya purpurascens (R. Br.) Bunge in Ledeb. Fl. Ross. 1: 195. 1841. Platypetalum purpurascens R. Br. in Parry’s Ist Voy. App. 267. 1823. Platypetalum dubiuin R. Br. ibid. Braya Thorild-Wul ffi Ostenf. in Medd. om Grdgnl. 64: 176. 1923. 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 47 ? Braya americana (Hook.) Fern. in Rhodora 28: 203. 1926. ? Braya Longtt Fern. ibid., 202. 1926, Braya purpurascens is a variable species which apparently is circumpolar. Where growing together with B. pilosa and B. glabella, it flowers at least three weeks later. The scapes are shorter and ascending, in fruit seldom more than 10 cm. high; the fruiting raceme is capitate or at least not much elongated. The mature siliques are plump, rarely 10 mm. long, and from 1.2 to 2 mm. thick. The style is short and thick, about 0.5 mm. long. The seeds apparently are sticky when wet but are not conspicuously mucilaginous. In North America B. purpurascens is distributed from the shores of Bering Sea along the Arctic Coast to eastern Greenland, north to the end of land and south to lat. 60° in Labrador (? Newfoundland) and to high mountains of Alberta. Apparently it is a decided calciphile. DROSERACEAE Drosera anglica Huds. Great Bear Lake: McTavish Arm, north shore, in a marl bog, 5237. New to the flora of the Northwest Territories. SAXIFRAGACEAE Saxifraga bronchialis L. ssp. Funstonii (Small) Hultén. RicHarpson Mrs.: West of Mackenzie Delta, 6684, 6829, 6830, New to the flora of the Northwest Territories. Saxifraga ferruginea Grah. MACKENZIE River Detta: Peel River, 1888, McConnell 8417 (Can), distributed as S. stellaris L. New to the flora of the Northwest Territories. Saxifraga flagellaris Willd. RicHarpson Mrs.: West of Mackenzie Delta, 6683, 6828. New to the flora of the Mackenzie District. Saxifraga foliolosa R. Br. Arctic Coast: Atkinson Pt., 2629. New to the flora of the Mackenzie District. Saxifraga nivalis L. GREAT BEaR LAKE: Dease Arm, 4842; McTavish Arm, 3783, 5204, 5307. The above numbers represent a new western limit in the Northwest Terri- tories, since previous records should be referred to S. reflexa Hook. Saxifraga reflexa Hook. The writer, in Rhodora 41: 242. 1939, has pointed out that the plant which in Alaska has passed for S. nivalis really belongs here (see notes on that species). In the Mackenzie Delta and on the Arctic Coast, S. reflexa is common east to Cape Dalhousie but apparently is not found far inland. Parnassia parviflora DC. GREAT BEAR LAKE: McTavish Arm, east shore, 3749. New to the flora of the Northwest Territories and not previously collected in the Mackenzie basin outside the Rocky Mountains. Ribes hudsonianum Richards. MacKenzie River Detta: Aklavik, 7309. 48 SARGENTIA \4 Common at Great Bear Lake and in rich woods north to some distance south of the tree limit. Ribes lacustre (Pers.) Poir. Mackenzie Distr.: Yellowknife R., Gordon Lake, J. Carroll 67A (Can). Ribes lacustre of Hook. Fl. Bor.-Am. 1: 232. 1834, “To Fort Franklin and Bear Lake, near the Arctic Circle,” is R. oxyacanthoides L., and the above is the first record of true /. lacustre from the Northwest Territories. ROSACEAE Spiraea Beauverdiana Schneid. See Rhodora 41: 244. 1939. This western species is very common in the higher parts of the Mackenzie Delta in heath and open birch woods, and near the East Branch it reaches the Arctic Coast. It extends but a short distance eastward, to the Eskimo Lake Basin, 2nd lake, 29917. Luetkea pectinata (Pursh) Kuntze. In the National Herbarium of Canada is a specimen labelled: N. W. T., Peel R., Mackenzie, 1889, McConnel 5722. McConnel crossed from the Mackenzie Delta into Yukon and Alaska via the Rat-Porcupine R. Pass, and this alpine species was no doubt collected in the mountains west of the Delta. Amelanchier sanguinea (Pursh) DC. Great Bear Lake: Bear River, lower slopes of Mt. Charles, 3298. Following Wiegand, in Rhodora 14: 138. 1912, the above specimens are best placed here. This is the “saskatoon berry” of the Mackenzie, common at least to the Good Hope Ramparts. Rubus pubescens Raf. MacKENZIE River De_ta: East Branch, 6550. Previously known in the Mackenzie basin north to Simpson. Potentilla ? pectinata Raf. Arctic Coast: Atkinson Pt., 2635. This plant has 3- to 5-parted basal leaves, almost circular in outline, with much broader lobes than in typical P. pectinata. The leaves are thinly tomentose- pruinose beneath but not silky-villose. It may be P. glabrella Rydb., a species not too well-marked from P. pectinata, but differs from it by the almost orbicular leaves and the large flowers, which have broad, deeply notched petals. Potentilla pensylvanica L. Ricuarpson Mrs.: West of Mackenzie Delta, 6603, 6638, 6834; EsKximo LAKE BASIN: 3036. Previously known northward to Great Bear Lake. Potentilla pulchella R. Br. var. gracilicaulis var. nov. Caules floriferi 5—30 cm. alti, 3—-8-flori, e basi ascendente erecti, folia 1 vel raro 2 in parte inferiore ferentes, tenuissimi, rubiginosi, sericeo-villosi mox glabres- centes ; folia basalia longipetiolata, sub-bijugata, foliola terminalia longe petiolata, superne glabrata, subtus sparse sericeo-villosa; stipulae castaneae, glabrae. Flores parvi sunt, illosque speciei in mentem vocant. Arctic Coast: Atkinson Pt., 70° N., 131° 20’ W., top of sandy cutbank, Aug. 1-3, 1927, A. E. & R.T. Porsild 3632 (rypr); same place, 2633; Tuktuayaktoq, 7425; Cape Dalhousie, 2776-2778; Cape Bathurst, Can. Arct. Exp. 511; YuKon Territory: Herschel Isl., Can. Arct, Exp. 241, 241A. 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 49 Our plant is a characteristic species in sandy places along the arctic coast east of the Mackenzie. It is abundantly distinct from the typical form except in the floral parts. It may be P. pulchella B elatior Lange, Consp. Fl. Groenl. 4. 1880, which is P. subarctica Rydb., in N. Am. Flora 22: 347. 1908, but if so it is very different from a plant from Hudson Bay which Rydberg, in the National Her- barium of Canada, has named P. subarctica. Polunin, |. c., has taken up var. elatior Lange for this, suggesting that it is a “luxuriant southern form.” Of B elatior Lange, Th. Wolf (in Bibl. Bot. 16| Heft 71]: 152. 1908) merely says: “. 2. die ich mit Abromeit (Bibl. bot. Heft 42. 7.) nur als ‘Standortsformen, die in einander ubergehen,’ auffasse. . . .” Potentilla pulchella var. gracilicaulis is, perhaps, less “arctic” than the species. In the large series collected at Cape Dalhousie (2766-2768) some specimens are only a few centimeters high, but none of them are transitional to the species. On Aug. 7-14, 1927, the plant was still in anthesis; it probably does not mature fruits there except in favorable seasons. Typical P. pulchella apparently is an eastern species which is rare or restricted to high latitudes west of Hudson Bay, and the plant which Hooker, |. c., reports from “Shores of the mainland between the Coppermine and Mackenzie Rivers. Dr. Richardson” is undoubtedly our plant. Potentilla rubricaulis Lehm. Nov. Stirp. Pug. 2: 11. 1830, et Rev. Pot. 68, tab. 30. 1856. Arctic Coast: Cape Dalhousie, 70° 20’ N., 129° 55’ W., dry sandy plain, Aug. 7-14, 1927, 2774, 2774A; Liverpool Bay, 2905. This is one of the rarest of the arctic Potentillae and one which, as pointed out by Simmons (FI. Ellesmere Land, 50-54. 1906), has been generally mis- understood by Rydberg and others. Because the type came from Bear Lake (‘About Bear Lake, in lat. 66°. Dr. Richardson”), the writer was particularly anxious to obtain authentic material. In this he failed in the Bear Lake Region, but on the Arctic Coast, at Cape Dalhousie, he obtained a large series of speci- mens that perfectly match Lehmann’s description and beautiful plate. Our specimens had immature fruits in the middle of August and also fully expanded flowers on the same plants. The flowers are almost 2 cm. in diameter, somewhat larger than in the type (Th. Wolf, 1. ¢., 170). Lehmann and Wolf have both remarked that P. rubricaulis bears some resemblance to 5-nate forms of P. nivea, but Wolf, 1. c., correctly states that it may at once be distinguished from P. nivea and forms by its conical style; it should be noted, however, that in the Mackenzie District the base of the style, in all forms of P. nivea examined by the writer, is enlarged and papillose, although the style itself is slender and filiform as described by Wolf. The tomentum of the undersides of the leaves, in our plant as well as in all other material of P. rubricaulis seen by the writer, is thinner than in most forms of P. nivea, so that the main nerves are distinctly visible. Also characteristic of P. rubricaulis is the fact that the tomentum extends in a small, white tuft, from each tooth or lobe of the leaflets. Distribution: Potentilla rubricaulis extends from Great Bear Lake and the Arctic Coast across the Arctic Archipelago east to N. W. Greenland and to the Franz Joseph Fiord region of the east coast. Sibbaldia procumbens L. RicHARDSON Mrs.: West of Mackenzie Delta, common locally in herb mats, 1500’ elev., 6833; Great Bear Lake: Etacho Pt., 1500’ elev., 3502; Keewatin Distr.: Kazan R., 5827, 6045; Lake on Tha-anne R., 5589. 50 SARGENTIA |4 The North American distribution of Sibbaldia is rather peculiar and very disrupted. It is common in E. & W. Greenland, north to 73° and 72°, rare on the Labrador coast, isolated in mountains of Newfoundland, Gaspé, and New Hampshire, fairly common in the Hudson Bay region south to James Bay, north to southern Baffin Island, and west to Central Keewatin. It is common in the high mountains of Alberta and British Columbia, Montana, Idaho, Wyoming, Colorado, and Utah, south to California, with two isolated, alpine stations in the Mackenzie District and one in Central Alaska. It is also common on islands of the Bering Sea. The species was previousy unrecorded from the Mackenzie District. Geum glaciale Adams. RicHarpson Mrs.: West of Mackenzie Delta, between 2000 and 4000’ elev., 6602, 0694. In Canada otherwise known only from Yukon Territory. New to the flora of the District of Mackenzie. Dryas octopetala L. MackKENzIE River De_ta: E. Branch, 7032; Campbell Lake, 1999; Richards Isl., 2098. A western species common in the mountains of Alaska and Yukon east to the Mackenzie Delta. East of the Delta it has been collected but a few times. LEGUMINOSAE Astragalus Collieri (Rydb.) comb. nov. Atelophragma Collieri Rydb. in Bull. Torr. Bot. Cl. 55: 127. 1928. Great Bear Lake: East shore of McTavish Arm, open spruce woods on sandy soil, 375/. Our specimens are with mature fruits; they match the description as well as the type, which came from Eagle, Alaska, June 29, 1902, A. C. Collier, 50 (US no. 379736), and also: Near Post on Forty Mile Creek, Yukon Valley, June 9, 1893, F. Funston 67, teste Rydb. (G), except that the pods in ours (collected late in the season) are merely puberulent. Astragalus Collieri superficially resembles A. frigidus var. americanus (Hook. ) S. Wats., from which it may at once be distinguished by its dark purplish stems and peduncles, which are covered with a minute white pubescence, and by its purplish brown pods. These are 2 cm. long, turgid, with a very narrow stipe 0.5 to 0.6 cm. long, and contain from 5 to 8 seeds. The seeds are dark olive green, dull, orbicular, about 2 mm. in diameter. The known distribution is: Eastern central Alaska, south to Glacier Bay, east to Great Bear Lake. New to the flora of Canada. Astragalus linearis (Rydb.) A. E. Porsild. See Rhodora 41: 250. 1939. Ricuarpson Mrs.: West of Mackenzie Delta, windswept gravelly ridges, 3000" elev., July 7-10, 1933, 6697 (flowering specimens) ; CoppERMINE River: Fort Hearne to Bloody Falls, 1931, A. M. Berry 14 (fruiting specimens) ; YUKON Territory: Dry Gulch, Gorman 1014; White Horse, Eastwood 625 [both in Can. Nat. Herb., teste Rydberg]. The above numbers appear to be the first collections from the Northwest Terri- tories, since a fragment in the National Herbarium of Canada from Great Bear Lake, collected by J. MW. Bell (22899, Can) and named Atelophragma lineare by Rydberg, is really Astragalus eucosmus Robins. Although Rydberg, in N. Am. Flora 24: 368. 1929, recorded A. lineare only from ‘Mackenzie and Yukon | Territories,” there is in the National Herbarium of Canada a typical specimen from Manitoba: Mouth of Qu’Appelle River, June 28, 1906, John Macoun & Wiliam Herriot 70479, labelled by Rydberg himself Atelophragma lineare. The 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 51 known distribution of Astragalus linearis is: Alaska, Yukon, Northwest Terri- tories east to Coppermine River and south to Manitoba. Astragalus yukonis M. E. Jones, Rev. Astr. 83, tab. 7, fig. 28. 1923. MAcKENzIE River Detta: South end of Richards Isl., on sandy riverbanks, where it formed large clumps, 21035. New to the flora of the Northwest Territories. Oxytropis arctica R. Br. in Suppl. App. Parry’s Ist Voy. 278. 1824; Hook. Fl. Bor.-Am. 1: 146. 1834 (var. subumbellata and var. microphylla) ; Gray in Proc. Am. Acad. 20: 4. 1884, in part; Simmons, Phytogeogr. Arct. Am. Arch. 112. 1913; Fern. in Rhodora 30: 144, tab. 172 (upper). 1928. Oxytropis Roaldi sensu Macoun & Holm in Rept. Can. Arct. Exp. 5A: 17, tab. 8, fig. 2. 1921, ? non Ostenfeld. Oxytropis coronaminis Fern. in Rhodora 30: 151, tab. 175. 1928. Of the numerous critical species of plants in the flora of boreal and arctic North America, most of the members of the genus Oxytropis, as shown by Fer- nald’s revision in Rhodora 30: 137-155. 1928, have been much misunderstood by earlier writers. Hooker, 1. c., clearly did not understand the genus, nor did John Macoun in his Catalogue of Canadian Plants, nor in more recent times J. M. Macoun and Theodore Holm. In addition to taxonomic difficulties, the genus offers a number of geographical puzzles, as pointed out by Simmons, I. c., 11}. In the course of many seasons’ collecting in many parts of the Canadian Arctic, from the Alaska boundary to Hudson Bay, the writer paid special attention to this genus. Of its purple-flowered members, two of the most elusive and puz- zling species appear to be O. Roaldi Ostenf. and O. arctica R. Br. In six sea- sons’ collecting in the Mackenzie District the writer never succeeded in finding the former, which thus far is known only from the type locality (Herschel Is- land). Of the latter Fernald, 1. c., in addition to the type (Melville Island), recognized but one collection (between Coppermine River and Cape Alexander, Rae) ; Ostenfeld, |. c., did not find it in the Gjoa Expedition collections. Several collections made by members of the Canadian Arctic Expedition, 1915-1918, and recorded by Macoun & Holm, |. c., as O. Roaldi, Fernald was unable to identify with either that species or with O. arctica. He therefore described the Can. Arct. Exp. plant as O. coronaminis, to be distinguished from O. arctica “by its much larger flowers, longer calyx-tube and -teeth and very large corolla with vexillum fully twice as broad as in true O. arctica.’ But in several numbers cited by Fernald in the National Herbarium of Canada, the size of the flower and the length of the corolla, calyx-tube and -teeth vary greatly. In McMillan’s collection from Melville Island (77294, Can), the corolla is 1.5 cm. long and the teeth of the calyx 4 as long as the tube. In a specimen from near the type locality of O. coronaminis (A. E. & R. T. Porsild 4745), the flowers are even smaller and the calyx-teeth vary in length from 4 to ¥4 of the length of the calyx- tube. Likewise in the stipules, often of systematic value in this genus, there does not appear to be any difference between O. coronaminis and O. arctica. To the description given by Fernald, |. c., for O. coronaminis, it may be added that the veins of the free blades of the stipules are green and very conspicuous; the main vein has a number of upward pointed side branches and ends in the green tip of the blade; the edges, besides the long, white ciliae, are beset with large, sessile glands or papillae. 2 SARGENTTA |4 Thus if O. coronaminis is conspecific with O. arctica, the distribution of the latter becomes more natural, making it an endemic of the central and western Arctic Archipelago and the mainland from Anderson River east to Bathurst Inlet, possibly with an isolated station in mountains of east central Alaska (Steese Highway, Eagle Summit, 3880’ elev., E. Scamman 806; Richardson Highway, Miller House, 4810’ elev., idem 807; both in Gray Herb.). The following speci- mens are in the writer’s collection from the Mackenzie District: Arctic Coast: Cape Dalhousie, old dunes and dry tundra, 2780, dry sandy slope, 2779B; Great Bear Lake: North shore of Dease Arm, near headwaters of Horton R., 4745. Oxytropis Belli (Britt.) Palibine. See Fernald in Rhodora 30: 150. 1928. To the description and discussion by Fernald, |. c., it may be added that QO. Belli forms large cushions a foot or more in diameter. The central root in old specimens sends out numerous freely forking branches, but these do not produce adventitious roots. In life the calyx is dark red with almost black teeth; the corolla is pinkish violet with a pale spot on the front of the vexillum and 4 or 5 bluish lines on each side; the scapes are reddish brown. On the west coast of Hudson Bay, O. Belli flowers before all other species of Oxytropis. The flowers are faintly fragrant. Oxytropis campestris (L.) DC. (sensu str.); Hook. Fl. Bor.-Am. 1: 147. 1834, var. melanocephala (in part only). Oxytropis glabrata Rydb. Fl. Prairies and Plains 485. 1932, non O. glabrata (Hook.) Nels. in Univ. Wyom. Publ. Bot. 1: 117. 1926. Acaulescent with a strong tap-root ending in a multicipital crown; stipules membranaceous, translucent, white, sparingy villous to glabrate, the free blades long-ciliate and with a few large papillae, subulate, prominently nerved; leaves 10-15 cm. long, leaflets opposite, lanceolate, acuminate, sparingly adpressed- sericeous, soon glabrate, with involute margins, dark green; scapes 15-20 cm. high, brownish green, sparingy silky villous, soon glabrate, erect or ascending ; spike subcapitate, slightly elongated in fruit, 10-15-flowered, the bracts linear, “4 as long as the calyx; calyx dark green with rather short grayish black hairs ; calyx-lobes black-hirsute, subulate, 4 as long as the tube; corolla about 1 cm. long, in life pale yellow or dirty white, somewhat darker when dry ; legumes gray- ish green, long-beaked, incompletely 2-celled, sparingly covered with short black hairs; seeds oblique-heartshaped, 1.5—-2 mm. in diameter, greenish gray. In open, sandy, or gravelly places from northern Labrador west to Yukon, not north of the tree-line. Hupson Bay: Churchill, 58° 47’ N., 94° 14’ W., gravelly plain, July 3, 1930, 5480; same place, J. M. Macoun 79102 (Can); same place, G. Gardner 68 (Can); same place, N. Polunin 1949 (Can), distr. as O. monticola Gray; Lasrapor: Komaktorvik Fj., 59° 17’ N., 63° 45’ W., Vv. C. Wynne-Edwards 7120 (Can) ; MAcKENzIE River: Fort Good Hope, Miss E. Taylor 5353 (Can), as O, Lamberti Pursh; YuKon: Carmacks, A. Eastwood 572 (Can). The above numbers are such a close match for O. campestris of northern Eu- rope that, although a number of related American species from time to time have been recorded under this name, there does not appear any good reason why the latter, in a strict sense, should not be applied to this plant. Oxytropis campes- tris, like O. Maydelliana, thus becomes almost circumpolar. Our plant is closely related to O. gracilis (A. Nels.) Schum., from which, however, it may at once be distinguished by the dark pubescence of the calyx and legumes. From the only other white-stipuled, pale-flowered northern species, O. hyperborea, it is easily distinguished by the absence of verticillate leaflets, the very characteristic subulate free parts of the stipules, and by the less dense vestment of hairs. 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES i When Rydberg examined most of the material in the National Herbarium of Canada, he annotated several sheets of O. campestris as “O. glabrata (Hook. ) A. Nels.” There can be little doubt that while Rydberg well understood the species, he misapplied Nelson’s interpretations of O. glabrata. When that au- thor made his transfer he obviously had in mind Hooker’s Oxvytropis campestris var. glabrata (foliolis glabriusculis subsucculentis), but what Hooker meant by this is not at all clear; it might have been the glabrate form of O. Maydelliana, although the “subsucculent leaflets” strongly suggests O. /iudsonica. There can be no doubt, however, as to which plant Nelson actually thought of, because in his discussion, 1. c., he adds: “Besides the characters given by Hooker, the fol- lowing drawn from the sheet cited below (17. W. Gorman, 1365, Nogheling R.., Alaska, July 21, 1902) may be added: Sparsely strigose hairy on leaves and scapes; the leaflets often nearly or even quite glabrous above; the branches of the caudex dark-brown from the numerous dead glabrous petioles.” This clearly shows that Nelson’s O. glabrata is O. Maydelliana Trautv. Rydberg, |. c., does not mention the colour of the stipules, but from his annotations on sheets of the pale-stipuled O. campestris it is obvious that by O. glabrata (Hook.) Nelson he did not mean O. Maydelliana Trautv., which, in the Ottawa collection, he annotated O. melanocephala (Hook.) Rydb. General distribution: Arctic and subarctic alpine Europe; Siberia; subarctic North America (? not in Alaska) ; not in Iceland. Oxytropis hudsonica (Greene) Fern. in Rhodora 30: 142. 1928. Ricuarpson Mrs.: West of Mackenzie Delta, 1500-2000’ elev., 6608, 6700, 6838; Arctic Coast: Liverpool Bay, 2908; Great Bear Lake: Ft. Franklin. Common in Keewatin District. In life the scapes of O. hudsonica are erect, ascending or sometimes decum- bent, and conspicuously angled. The entire plant is fresh green and very viscid to the touch; the leaflets are somewhat fleshy and, due to the strongly involute margins, somewhat boat-shaped, obscurely keeled. The flowers are reddish purple, becoming almost blue in fading, and very fragrant with a perfume remi- niscent of roses. The mature seeds are coal-black, reniform, smooth, dull, 1.6— 1.8 x 1.2-1.4 mm. Oxytropis hyperborea sp. nov. Oxytropis campestris var. melanocephala Hook. Fl. Bor.-Am, 1: 147. 1834, in part. Oxytropis campestris (L.) DC. var. sordida Willd. sensu Holm & Macoun in Rept. Can. Arct. Exp. 5A: 16. 1924, in part. Oxytropis ? borealis sensu Fern. in Rhodora 30: 143. 1928, non DC. Prodr. 2: 275. 1825. Herba acaulescens radice crassa descendente; folia 3-12 cm. longa, stipulae membranaceae vel chartaceae pallidae vel albescentes, longe sericeo-villosae, per- sistentes, laminae parte libera longe ciliata, longe attenuata et haud subulata, prominenter nervosa; foliola lineari-lanceolata, 1 cm. longa, 0.3-0.4 cm. lata, subverticillata, 3- vel 4-juga, matura ob rhachem elongatum habitu interdum alter- nata vel opposita, utrinque cum rhachi sericeo-villosa, supra in speciminibus e locis udis interdum glabrata, tenella margine subrevoluta, adulta revoluta vel fere revoluta; scapi 5-20 cm. alti, longe sericeo-villosi, vulgo erecti sed in spe- ciminibus e regionibus borealibus ultimis plus minusve prostrati; spica sub- capitata, sublaxiflora, elongans, 6—-10-flora, floribus infimis saepius sat discretis ; bracteae lineares calycem aequantes ; calyx albus, sericeo-villosus, indumento pilis nigris abbreviatis commixto, quapropter, pubescentia albicante aetate labefacta, calyx obscuriore colore evadens; calycis lobi triangulares, apice subulati quam tubo 24 breviores; corolla circa 1.5 cm. longa, in speciminibus vivis ochroleuca 54 SARGENTTA \4 vel sordide albida, exsiccans in vexillo interdum puncto nigricante (in plantis vivis nullo vel vix obvio) maculata; legumina 1—1.5 cm. longa, subinflata, longiro- strata, imperfecte bilocularia, indumento denso sericeo-villoso, pilis nigris paucis brevibus intermixto demum evanido. In sandy and gravelly places or in dry tundra, from mountains of eastern Alaska and Yukon east along the Arctic Coast to Coronation Gulf, south to Great Bear Lake. MAcKENzIE River Detta: East Branch, lat. 68° 55’ N., dry sandy slopes, July 21, 1934, A, E, Porsild 7033 (tyre). In addition to the type the following specimens, all in the National Herbarium of Canada, may be considered typical: ALASKA: Near Head of Chitina R., dry hillsides, 2500’ alt.. H. M. Laing 134, 135; Arctic Coast oF YUKON TERRITORY : 69° 12’ N., 138° to 138° 30’ W., 7151; Mackenzie River Detta: South end of Richards Isl., dry slopes, 7072; same vicinity, 2106; north end of Richards Isl., 2253; same vicinity, 7449; Arctic Coast: 6 miles east of Kittigazuit, sand dunes, 2502; Cape Dalhousie, old dunes and dry tundra, 2779; Great Bear Lake: North shore of Dease Arm, dry tundra and sand dunes, #705; Cape McDonnel, low tundra and lake shore, 5155; Etacho Pt., elev. 1500’, alpine tundra, 3503; Leith Pt., open woods and sandy beaches, 3583; McTavish Arm, Hunter Bay, shady ravines near cold spring, 5328; DotpH1n & Union Srrait: Bernard Harbour, Can. Arct. Exp, 332 (in part); Clifton Pt, Can. Arct. Exp. 690, 691; Victoria Isl., Can. Arct. Exp. 386. Oxytropis hyperborea is well-marked from the other yellow-flowered northern species by its verticillate leaflets, and from O. Maydelliana also by its pale stipules (for contrast with O. campestris see treatment of this species). It belongs in the group of arctic plants which Hooker and most later authors have cited under O. campestris, O. campestris vars. glabrata, sordida and melanocephala, O. leu- cantha, and O. monticola. Fernald, |. c., 142-144, when he wrote about this group and showed that the castaneous- or fulvous-stipuled plant is really O. May- delliana, perhaps was the first to point to the existence of still another clear-cut species which he tentatively identified with O. borealis DC., from extreme E. Siberia. Because de Candolle listed O. borealis under the heading “*floribus purpurascentibus aut albis” rather than under “**floribus ochroleucis,” and in view of the following passages in his description: “foliolis . . . subtus glabris . scapis folii longitudine, floribus capitatis, bracteis calycis nigro-hispidissimi longitudine,” it seems scarcely possible to identify our Canadian arctic plant with O. borealis DC. This view is strengthened by the fact that O. hyperborea does not seem to occur in N. W. Alaska, nor has the writer seen any other pale- stipuled, yellow-flowered Oxytropis from northern Alaska. The fact that Hooker (Fl. Bor.-Am, 1: 145) cited specimens of O. borealis collected by Lay and Collie (Kotzebue Sd.) 1s no proof of its presence in Alaska. These specimens may well have been O. Maydelliana, because Hooker obviously did not understand the northern members of this genus well. It may be significant also that Hooker, l. c., when copying (otherwise verbatim) de Candolle’s description, left out the last epithet : “Stipulae pallidae.” For his treatment of Ovytropis for “North American Flora,” Rydberg exam- ined all arctic material in the National Herbarium of Canada. The low, north- ern form of O. hyperborea as exemplified by Can. Arct. Exp. 332 from Bernard Harbour, Aug. 15, F’. Johansen, and nos. 690 and 691 from Clifton Point, is an- notated by Rydberg as “Oxytropis Johansenti Rydb.” Unfortunately sheet no. 332, marked “type” by Rydberg, contains three plants, two of which are O. hyperborea, while the third is typical O. Maydelliana. The isotype of “O. Johansenii,” in N. Y. Bot. Gard., is all O. hyperborea. Rydberg’s “O. Johan- senit” (not to be confused with O. Johannensis Fern.) was never published. Oxytropis monticola Gray and O. alpicola Rydb. both have pale stipules. The on e 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES first has purple flowers and the latter is said to have decumbent peduncles and an ochroleucous corolla with a purple spot on the keel. Oxytropis pygmaea (Pall.) Fern. RicHarpson Mrts.: West of Mackenzie Delta, 2000-3500’ elev., 6699, 6841; MACKENZIE River Detta: East Branch, 7037. New to the flora of the Northwest Territories. Oxytropis retrorsa Fern. Eskimo Lake Basin: Ist lake, 6936; Great BEAR Lake: Bear River, 3365. New to the flora of the Northwest Territories. Lathyrus ? palustris L. A Lathyrus, probably of this species, was observed in winter along the dog trail above Arctic Red River on the lower Mackenzie. CALLITRICHACEAE Callitriche autumnalis L. MACKENZIE River Detta: East Branch, 6920, 7237; Eskimo LAKE BasIN: 2995, 3063, 6934; Great Bear Lake: Bear River, 3366; Keith Arm, 3438; McTavish Arm, 3697. Perhaps common or at least occasional north to the limit of trees. Callitriche verna L. emend. Lonnroth. Eskimo LAKE Basin: Setidgi Lake, 3180; Great Bear Lake: McTavish Arm, 3713, 3697; Keith Arm, 5279; Krewatin Distr.: Lake on Tha-anne R., 6179; Lower Kazan R., 6058, 6059. Probably common throughout the Northwest Territories north to the limit of trees or slightly beyond. VIOLACEAE Viola labradorica Schrank. Great BEAR LAKE: South shore, Leith Pt., 3546; McTavish Arm, 3584. New to the flora of the Mackenzie District. Viola Langsdorffii I isch. RicHarpson Mrs.: West of Mackenzie Delta, 7342. New to the flora of the Northwest Territories. Viola pallens (Banks) Brainerd. GREAT BEAR LAKE: Etacho Pt., 1500’ elev., 3504; Kerewatin Distr.: Kazan R., 5831, 5832, 6056; Lake on Tha-anne R., 5591. New to the flora of the Northwest Territories. Viola renifolia Gray var. Brainerdii (Greene) Fern. GreAT BeaR LAKE: McTavish Arm, 53/2. ELAEAGNACEAE Shepherdia canadensis (L.) Nutt. MAcKENZIE River Detta: Campbell Lake, 1952, 2004; Richards Isl., 27/1; Eskimo Lake BAsIN: 3043, 3160; GreEAT Bear LAKE: Ft. Franklin, 3220; Dease Arm, 4776, 4814. Common or occasional on calcareous soil north to the Mackenzie Delta but not observed east of the Palaeozoic formation. At its northern limit all the plants appeared to be male. ONAGRACEAE Epilobium arcticum Samuelss. RicHARDSON Mrs.: West of Mackenzie Delta, 6704, 6842. Previously known in Canada only from the Eastern Arctic. 56 SARGENTIA \4 Epilobium davuricum [isch. Arctic Coast oF YUKON Territory: King Pt., 7187; MAcKENzIE River DeLtta: Camp- bell Lake, 1939; Kittigazuit Isl., 2376, 2506; Great Bear LAKE: Keith Arm, Deerpass Bay, 3445; Dease Arm, 4846; KEEWATIN Distr.: Kazan R., 5833, 6054. New to the flora of the Mackenzie District. Epilobium palustre L. Common northward beyond the limit of trees, and represented by many num- bers in the writer's collections from the Mackenzie Delta, the Arctic Coast, Great Bear Lake, and Central Keewatin. Previously recorded in the Mackenzie basin north to 64° (Richardson). HIPPURIDACEAE Hippuris tetraphylla L. f. See Rhodora 41: 264. 1939, Arctic Coast oF YUKON TERRITORY: Shingle Pt., 6907. New to the flora of Canada. HALORRHAGACEAE Myriophyllum alterniflorum DC. Eskimo Lake Basin: North shore of 2nd lake, 2997; Great Bear LAKE: Etacho Pt., 1500’ elev., 3506; Dease Arm, 4706. Our specimens are all sterile. New to the flora of the Northwest Territories. Myriophyllum exalbescens I’ern. MackenzigE River Detta: Aklavik, 7308; Campbell Lake, 2026; East Branch, 6927, 7239, 7249; Eskimo Lake Basin: 2nd lake, 2998; Setidgi Lake, 31/1. Myriophyllum exalbescens, as elsewhere, appears to be more northern than M. alterniflorum, and in the Mackenzie Delta it may normally produce mature fruit. UMBELLIFERAE Bupleurum americanum Coult. & Rose. Ricnarpson Mrs.: West of Mackenzie Delta, to 4000’ elev., 6705, 6844; MACKENZIE River Detta: East Branch, 7041; Campbell Lake, 1942; Richards Isl., 2108. A western species which has merely crossed the Mackenzie, not previously recorded from the Northwest Territories. Cicuta mackenzieana Raup in Jour. Arn. Arb. 17: 279, pl. 197. 1936. Eskimo LAKE Basin: Setidgi Lake, 3177. Not previously known in the district north of Great Slave Lake (Raup). Conioselinum cnidiifolium (Turcz.) A. E. Porsild. See Rhodora 41: 267. 1939. This plant belongs in that group of western species that reach their eastern limit near the Mackenzie Delta, but, unlike most of these, it has penetrated a few hundred miles along the Arctic Coast to the mouth of the Anderson River (Liver- pool Bay, 2974). It is very common on sandy river banks in the Delta and is probably the plant which Richardson (Arctic Searching Exp. 1: 233. 1851) reports under “Musenion divaricatum, East Branch, in lat. 68° 25’ on sandy banks.” CORNACEAE Cornus canadensis L. Great Bear Lake: Bear River, Mt. Charles, 3317, 3350, 3351. Hooker’s statement, I. c., “Throughout Canada, nearly to the Arctic Coast, everywhere as far as the Pine woods extend.—Kichardson,” certainly needs veri- eal = 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES fication, especially as “Pine woods” here probably is intended to mean “‘conifer- ous woods.” In the writer’s experience, C. canadensis does not extend beyond the Canadian zone, and in the Mackenzie District it is limited to the Mackenzie Valley and Great Slave Lake. Cornus stolonifera Michx. Great Bear LAKE: Bear River, Mt. Charles, forming thickets by a small stream, 3339. Unrecorded from the Northwest Territories, although specimens from Simp- son and Norman, collected by Miss E. Taylor, are in the National Herbarium of Canada. Hooker’s report, under C. alba L., “north to lat. 69°—Richardson,” is certainly erroneous and is probably a misprint for “59°.” Richardson, in Franklin’s Journ. App. 732, says ‘“‘W” (wooded country between lat. 54 and 64). PY ROA GH Aas Pyrola chlorantha Sw. MAcKENzIE River: Sans Sault Rapids, 6570; Bear Rock, Ft. Norman, 3389; Great BEAR LAKE: Bear R., 3286; Mt. Charles, 3335; McTavish Arm, north shore, 53/4; Leith Pt., 3549. Rare or occasional in cold spruce woods, north in the Mackenzie basin to Ft. Norman and Bear Lake. Pyrola minor L., Hook. Fl. Bor.-Am. 2: 45. 1834; Simmons, Phytogeogr. Arct. Am. Arch. 115. 1913. Ricuarpson Mrs.: West of Mackenzie Delta, 7343; Mackenzie River Deta: East Branch, 6552; Great Bear Lake: Etacho Pt., 1500’ elev., 3508; Leith Pt., 3585, 3586. A rare species of peculiar and disrupted range. Hooker's statement, I. c., “Barren country from lat. 64° to the Arctic Coast and Islands,” certainly is erroneous. Very questionable also is the record from Melville Peninsula, Sim- mons, Il. c. The above are the first authentic collections in the Northwest Territories. Pyrola secunda L. MACKENZIE River: Sans Sault Rapids, 6569. Pyrola secunda L. var. obtusata Turcz. See Porsild in Rhodora 41: 274, 1939. Ricuarpson Mrts.: West of Mackenzie Delta, 6611; MackKENzIE River DELTA: East Branch, 6966; Campbell Lake, 1943, 1944; Richards Isl., 2262; Kittigazuit Isl., 2387; Arctic Coast: 6 mi. east of Kittigazuit, 2507; Atkinson Pt., 2643; Liverpool Bay, 2915; Eskimo Laxe Basin: 3000, 3044; Great Bear Lake: Ft. Franklin, 3223; Bear R., 3319; Cape McDonnel, 3/56; Dease Arm, Narakay Isl., 4848; McTavish Arm, north shore, 53/5; south shore, 3633; KeEwaTIN Distr.: Yathkyed Lake on Kazan R., 5838; Lower Kazan R., 6060; west coast of Hudson Bay, 5704. A species characteristic of northern coniferous woods but which, in the conti- nental Northwest Territories, reaches far north of the limit of trees. ERICACEAE Loiseleuria procumbens (L.) Desv. Ricuarpson Mrs.: West of Mackenzie Delta, June 20, 1931, O. Bryant; Great BEAR LAKE: Dease Arm, north shore, #707; Krewatin Distr.: Baker Lake, 6102; Lake on Tha- anne R., 5594. A rare species in the Mackenzie District, limited to acid pre-Cambrian rocks. Kalmia polifolia Wang. Great Bear LAKE: Keith Arm, 3429; Etacho Pt., 1500’ elev., 3509; KrEwatin Distr. : Lake on Tha-anne R., 5595. 58 SARGENTIA \4 A plant of northern muskegs that in Keewatin extends some distance beyond the limit of trees; rare in the lower Mackenzie Basin, north to Bear Lake and in Central Yukon. Phyllodoce coerulea (L.) Bab. KEEWATIN Distr.: Thelon R., J. B. Tyrrell 15563, 23194 (Can); Thelon Sanctuary, C. H. D. Clarke 808. An eastern species which reaches west to Central Keewatin. Cassiope hypnoides (L.) D. Don. Keewatin Distr.: Thelon Sanctuary, C. H. D. Clarke 804; Maguse R., W. Giissow 57. Like the preceding, an eastern species which reaches but a short distance west of Hudson Bay. Chamaedaphne calyculata (L.) Moench. Ricnarpson Mrs.: West of Mackenzie Delta, 6614; Mackenzie River Decta: Aklavik, 7311; Campbell Lake, 1953; Eskimo Lake Basin: 3006, 3047; Setidgi L., 3114; Great BeaR LAKE: McTavish Arm, 5249. Rare in the northern parts of the district, barely reaching to the edge of the timber. Arctostaphylos alpina (L.) Spreng. MACKENZIE River Detta: East Branch, 7276; Kittigazuit Isl., 23854; Arctic Coast: Atkinson Pt. 2645; Eskimo Lake Basin: 3005; Great Bear LAKE: McTavish Arm, 5246; KEEwaTIN Distr.: Seton & Preble 78353 (Can): Thelon Sanctuary, C. H. D. Clarke 817, 818. Occasional in acid sandy or gravelly places, chiefly on the pre-Cambrian rocks. Arctostaphylos rubra (Rehd. & Wils.) Fern. Very common on the Arctic Coast from Herschel Island east to Coronation Gulf and south to Great Bear and Great Slave Lakes, apparently preferring cal- careous soil and for this reason best developed on the Palaeozoic rocks. The plant recorded by Simmons, I. c., 118, under A. alpina, from Banks and Victoria Is- lands, probably belongs here. Oxycoccus microcarpus Turcz. See Porsild in Can. Field Nat. 52: 116-117. 1938. MackKENZIE River De_ta: Campbell L., 1949; Eskimo Lake Basin: 3004; GREAT BEAR LAKE: North shore, 5047; Smith Arm, 5013; McTavish Arm, north shore, 5245; south shore, 3753; Keewatin Distr.: Yathkyed Lake on Kazan R., 5839, Common in sphagnum bogs of the forested country, extending but a short distance beyond the limit of trees. DIAPENSIACEAE Diapensia lapponica L. KEEWATIN Distr.: Yathkyed Lake on Kazan R., 5844; Casba Lake, Tyrrell 23114 (Can). An eastern species which reaches but a short distance west of Hudson Bay. Diapensia obovata (Fr. Schm.) Nakai. See Porsild in Trans. Royal Soc. Can. ser. 3. sect. 5, 32: 35. 1938. RicHarpson Mrts.: West of Mackenzie Delta, barren granite scree, 3000’ elev., 6708; lower slopes, elev. 1500’, 6845, 7344. In the National Herbarium of Canada are several collections from central Yukon, but so far the species has not been collected east of the Mackenzie. PRIMULACEAE Primula borealis Duby. See Fernald in Rhodora 30: 94. 1928. Great Bear Lake: Dease Arm, north shore, 4713; McTavish Arm, north shore, 5/80. 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 59 A western species common along the Arctic Coast eastward to Cape Bathurst ; apparently rare in the interior. Primula incana M. E. Jones. See Fernald in Rhodora 30: 72. 1928. MACKENZIE River: Bear Rock, near Ft. Norman, 3390. Not previously recorded from north of Great Slave Lake. Douglasia arctica Hook. Fl. Bor.-Am. 2: 120. 1838; Ostenf. in Gjoa Exp. 60. 1909; Holm & Macoun in Rept. Can. Arct. Exp. 5A: 18. 1921; Constance in Am. Mid. Nat. 19: 258. 1938. Arctic Coast oF YUKON Territory: Shingle Pt., 6730; between King and Kay Pts., 7193; Ricuarpson Mrs.: West of Mackenzie Delta, between 3000 and 4000’ elev., 6617, 6709, 6710, 6846. Hooker’s statement, I. c., “Arctic sea-shore, between the Mackenzie and Cop- permine Rivers. Dr. Richardson,” needs confirmation, since the species has not since been collected east of the Mackenzie, whereas it is quite common on the Arctic Coast to the westward and on alpine slopes of the Richardson Mountains, where it is found only on snow-flushes. The flowers appear shortly after the snow melts and during anthesis are almost hidden among the leaves. Following anthesis the peduncles rapidly elongate to about 2cm. When past flowering, the plant is very inconspicuous and, due to the reddish brown color of the marces- cent leaves, looks like a tuft of “moss.” The flowers vary considerably in size and color, from 0.5 to 1 cm. in diam., and from rose-pink to deep purple. The seeds are large (1.8 & 1.3 mm.), dull, black and wrinkled, two in each capsule. GENTIANACEAE Gentiana Amarella L. Great Bear Lake: Bear R., 3369, 3370; Etacho Pt., 1500’ elev., 3510; Leith Pt., south shore, 3552, A southern species, previously collected but once north of Great Slave Lake. Gentiana arctophila Griseb. in Hook. Fl. Bor.-Am. 2: 61, tab. 149. 1838; Holm & Macoun, Rept. Can. Arct. Exp. 5A: 19. 1921, in part. Arctic Coast: Liverpool Bay, 2927; Eskimo Lake Basin: 3072, 3073; GREAT BEAR Lake: MeTavish Arm, north shore. The above-cited specimens are a good match for Grisebach’s description and plate. Gentiana arctophila is no doubt closely related to the widely distributed G. propinqua, but it is a much lower plant; well-developed specimens as a rule are freely branched from the base, with strongly ascending branches. The ter- minal flower is much larger than the lateral ones. The corolla is funnel-shaped ; the lobes are elliptic, distinctly mucronate, half as long as the tube in the terminal flower while equaling it in the lateral ones. The seeds average 0.8 < 0.6 mm. and are short-elliptic, round, not flattened in cross-section, light brown, smooth, but not shiny. Gentiana ? detonsa Rottb., Griseb. in Hook. Fl. Bor.-Am. 2: 64. 1838, as regards plant of “Arctic Sea-shore” (as G. detonsa [Fries]). For discussion of the nomenclature, history, and identity of this plant see Morten P. Porsild in Medd. om Gr¢gnl, 93°: 43- Dan 2a Arctic Coast: Atkinson Pt., on a dry, sandy slope 8 miles from the coast, Aug. 1-3, 1927, 2653. Theo. Holm, who, in Ott. Nat. 15: 175-183. 1901, revised the American mem- bers of the Section Crossopetalae, denied the presence in America of G. detonsa (G. serrata Gunn.), without, however, attempting to determine what plants 60 SARGENTIA 14 Grisebach, |. c., had had in mind. East of the Mackenzie Delta the writer, in 1927, collected a series of specimens of a gentian that in the field he at once identified with the G. detonsa with which he was familiar in Greenland. When compared with specimens from Greenland and Norway, it becomes apparent that the series differs somewhat from typical G. detonsa. In the Mackenzie plant the lobes of the corolla taper to a narrow point, whereas in the Greenland plant they are oblong and rather obtuse; also, the basal leaves in our plant are some- what broader and more obtuse. Since the plant otherwise is a close match for typical G. detonsa, the writer, at least until further collections are made, prefers to retain it under this name. Gentiana Raupii sp. nov. Gentiana elegans sensu Raup in Nat. Mus. Can. Bull. 74: 157. 1935, in Jour. Arn. Arb. 17: 290. 1936, non Nels. Gentiana procera sensu Porsild in Rhodora 41: 279. 1939, non Holm. Herba annua glaberrima, caules tenues, obscure purpurascentes, angulati, sim- plices vel e basi nodisque ramosi, 2-60 cm. alti; folia basalia oblanceolata, brevi- petiolata; folia caulina opposita, 2- vel 3-juga, lineari-lanceolata, sessilia, sub- amplexicaulia; pedunculi longi, uniflori; calyx purpureo-viridis, fere ad medium fissus, lobi subaequales, lanceolati, acuminati, marginibus membranaceis, obscure carinatis glaberrimi; corolla tenuis, purpureo-cyanea, 1-4 cm. longa, 4-lobata, fere ad medium incisa, lobis oblongo-truncatis, margine fimbriatis, apice eroso- dentatis ; glandulae nectariferae praesentes; stamina 4, filamentis glabris, dilatis vel infra alatis; stylus liber, latitudinem stigmatis semilunaris bilobati aequans ; semina oblonga, papillis brevibus obtusis dense tecta. Mackenzie River: Near Blackwater R., 62° 63’, alluvial clay banks, Sept. 17, 1931, A. E, Porsild 6571 (type) ; ramparts below Good Hope, 3403; M. Jones, 1903, 62299 (Can) ; GREAT SLave Lake: C. F. Howe 91958 (Can): G. S. Hume 102678 (Can): Woop BuFFALo Px.: H. M. Raup 3026, 3028, distr. as G. elegans; ALASKA: Richardson Highway, between Summit and McCarthy, 4. E. & R. T. Porsild 440 (distr. as G. procera). Gentiana Raupii is closely related to G. procera, G. Macounii, and G. elegans, but it differs from all in being less leafy and in the absence of the very character- istic papillose keels of the calyx-lobes ; the style is longer than in the first two but much shorter than in G. elegans; from G. detonsa it may at once be distinguished by the fringed corolla-lobes. It appears to be common in the Mackenzie Basin in alluvial meadows and on clay banks. In the type locality flowering and fruiting specimens were col- lected together. Gentiana glauca Pall. Ricuarpson Mrs.: West of Mackenzie Delta, 6618, 7345. Not previously recorded from the Northwest Territories. Lomatogonium rotatum (L.) Fries. Pleurogyne rotata (L.) Griseb. Holm & Macoun in Rept. Can. Arct. Exp. 5A: 19, tab. XI, fig. 3. 1921. Pleurogyne Carinthiaca sensu Holm & Macoun, ibid., non Wulf. See Fernald in Rhodora 21: 197. 1919. Arctic Coast oF YuKON Territory: Shingle Pt., 7113; Arctic Coast: West of Mac- kenzie, Tuktuayaktoq, 7437; Atkinson Pt., 2654; Cape Dalhousie, 2789; Kugaruk, west of Anderson R., 3087; Great Bear Lake: McTavish Arm, north shore, vicinity of cold min- eral spring, 5/83. Our specimens from the Arctic Coast west of the Mackenzie match the plant which Holm and Macoun referred to P. Carinthiaca. It differs from typical L. 1943} PORSILD, FLORA OF THE NORTHWEST TERRITORIES 61 rotatum in its shorter and broader calyx-lobes, its slender and less strict growth, and in its smaller seeds; these are dark brown and average 0.4 X 0.2 mm. as compared with 0.6 < 0.3 mm. in typical L. rotatum (in P. Carinthiaca, accord- ing to Hegi, Fl. Mitteleur. 5°: 1977, 0.5-0.7 0.3-0.5 mm.). Our plant also is not halophilous but, like Gentiana propinqua and G. arctophila, grows in herb mats, often far from the sea-shore. In view of the polymorphism in L. rotatum, I am here, with Fernald, 1. c., maintaining the plant under this name. Menyanthes trifoliata L. MacKENzIE River Detta: Campbell Lake, 1956; Arctic Coast: East of Kittigazuit, 2514; Esximo Lake Basin: Setidgi L., 3117; Great Bear Lake: McTavish Arm, east shore, 3754, 5355; north shore, 5250; north shore, Haldane R., 5049; Krewatin Distr. : Yathkyed Lake on Kazan R., 5842. Somewhat rare in the wooded country, occasionally extending beyond the limit of trees. Inthe Mackenzie District M. trifoliata is a late-flowering species which matures fruit only in favorable seasons. Compare also Fernald, in Rhodora 31: 105. 1929, and Porsild, ibid. 41: 281. 1939. POLEMONIACEAE Phlox Richardsonii Hook. Fl. Bor.-Am. 2: 73, tab. 160. 1838; Holm & Macoun, Rept. Can. Arct. Exp. 5A: 19, 1921. Rricuarpson Mrs.: West of Mackenzie Delta, 2500-3000’ elev., 6712. Our plants grew on volcanic tufa and formed large, loose mats. The leaves are lanceolate, not subulate, and the flowers are large, 1.5 to 2 cm. in diameter. Polemonium acutiflorum Willd. See Hultén, Fl. Kamtch. 4: 72. 1930. Polemonium coeruleum L. var. villosum (Rud.) Brand, Ostenf. in Gjéa Exp. 62. 1909; Holm & Macoun in Rept. Can. Arct. Exp. 5A: 19. 1921. Ricuarpson Mrs.: West of Mackenzie Delta, 6619, 6713, 6848; Arctic Coast oF YUKON Territory: Shingle Pt., 6909, 7112; King Pt. 7194; Mackenzie River Detta: Richards Isl., 2276; Arctic Coast: Liverpool Bay, 2928. Polemonium acutiflorum is an Asiatic species which in western America reaches but a short distance east of the Mackenzie. It inhabits moist, fertile slopes and snow-flushes, where it flowers late and in some seasons does not ma- ture seed. Polemonium pulcherrimum Hook. in Bot. Mag. 57: tab. 2979. 1830; Brand in Pflanzenr. 27 (IV. 250): 34. 1907. Mackenzie River Derta: Campbell Lake, 1959; East Branch, 7043; Eskimo LAKE Basin : 6937; Great BEAR LAKE: Etacho Pt., 1500’ elev., 3511. Our plants match a large series in the National Herbarium of Canada from Yukon Territory, Alaska, and British Columbia, lying mostly under P. humiule. Although Hooker himself, in Fl. Bor.-Am., reduced P. pulcherrimum to an un- important variation of P. coeruleum, there can be little doubt that the plant is abundantly distinct. Polemonium pulcherrimum apparently is restricted to cal- careous soils. We collected it on ledges of limestone cliffs, where it flowers very early. The plant is very viscid and fetid of odor. The flowers are much smaller than in P. acutiflorum, while the seeds are quite wingless and more or less lustrous. The plant from Bear Lake differs somewhat by its larger flowers and by being less viscid. 62 SARGENTIA \4 BORAGINACEAE Lappula Redowskii (Hornem.) Greene var. occidentalis (Wats.) Rydb. Eskimo Lake Bastin: 3050. Previously known in the district northward to Great Bear Lake. Eritrichium aretioides (Cham. & Schlecht.) DC. Arctic Coast or Yukon Territory: Between King and Kay Pts., 7159. A rare plant, known in North America from a few places in arctic Alaska and Yukon Territory. See Porsild in Rhodora 41: 282. 1939, where first recorded from Canada. Mertensia paniculata (Ait.) G. Don. See L. O. Williams in Ann. Mo. Bot. Gard. 24: 42. 1937. Mertensia pilosa (Cham.) G. Don, Ostenf. in Gjja Exp. 63. 1909, Arctic Coast or Yukon Territory: Between King and Kay Pts., 7192, where previously collected by the Gjé6a Expedition. The record in Fl. Bor.-Am., under Lithospermum paniculatum, from Great Bear Lake, needs confirmation. LABIATAE Stachys palustris L. Mackenzie River: Lat. 62° 30’, 6564. Common in the upper Mackenzie basin and re- corded in Fl. Bor.-Am. from Great Bear Lake. Greene, Pitt. 3: 342. 1898, has proposed the name S. scopulorum for the west- ern plant. Rydberg and most recent writers have taken up that name, but the plant of Alaska and Mackenzie, at least, seems inseparable from the almost cir- cumpolar S. palustris L. SCROPHULARIACEAE Limosella aquatica L. GREAT Bear Lake: McTavish Arm, east shore, 5357; south shore, 3684, 3708. Our specimens grew in small, shallow ponds on granite hills and were in fruit the second week of August. The seeds are 0.6 to 0.7 mm. long and conspicu- ously transversely wrinkled. Although perhaps often overlooked, L. aquatica is a rare plant in Canada and has been collected but a few times from southern Labrador to British Columbia. Previously known north to Lake Athabaska and Churchill, Man. (A. E. Porsild 5499). Veronica alpina L. var. unalaschcensis Cham. & Schlecht. See Fernald in Rhodora 41: 447-457. 1939. Veronica Wormskjoldi Roem, & Schult. Great Bear Lake: Etacho Pt., 1500’ elev., 3512; Keewatin Distr.: Lake on Tha-anne R., 5600. New to the flora of the Mackenzie and Keewatin Districts. Lagotis Stelleri (Cham. & Schlecht.) Rupr. Gymnandra Stelleri (Cham. & Schlecht.) Hook. Fl. Bor.-Am. 2: 102, 1838. Lagotis glauca Gartn., Ostenf. in Gjda Exp. 66. 1909. Lagotis glauca Gaertn. var. Stelleri Cham. & Schlecht., Macoun & Holm in Rept. Can. Arct. Exp. 5A: 20. 1921. Arctic Coast oF YuKoN Territory: King and Kay Pts., 7198; RicHarpson Mrts.: West of Mackenzie Delta, wet alpine tundra, between 2000’ and 4000’ elev., 6715, 6850. A western species which barely reaches the Mackenzie River. See also Por- sild in Rhodora 41: 283-284. 1939, 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 63 Castilleja hyperborea Pennell in Proc. Acad. Nat. Sci. Phil. 86: 532. 1934; see also Porsild in Rhodora 41: 284, 1939. RicHarpson Mts.: West of Mackenzie Delta, dry, sandy ridges, 800’—2000’ elev., 6851; MacKENZIE River Detta: Campbell Lake, gravelly limestone plateau. Apparently an endemic species of Alaska and Yukon Territory, which has reached the eastern bank of the Mackenzie Delta. Castilleja pallida (L.) Spreng. ssp. caudata Pennell in Proc. Acad. Nat. Sci. Phil. 86: 524. 1934. Castilleja septentrionalis Hook. Fl. Bor.-Am. 2: 105. 1838, at least as regards plant of “Arctic Sea-coast.” Castilleja pallida (L.) H. B. K., Ostenf. in Gjoa Exp. 64. 1909; Simmons, Phytogeogr. Arct. Am, Arch. 122. 1913. See also Porsild in Rhodora 41; 285. 1939. Arctic Coast oF YUKON TERRITORY: Common between King and Kay Pts., 7160; MACKENZIE River DeL_ta: Husky R., 6623; Campbell L., 2010. The plant recorded from Banks and Victoria Islands by Simmons (as var. septentrionalis) probably belongs here. Apparently an endemic species of Alaska and Yukon Territory, which reaches east to the Mackenzie Delta and north to Banks Island. Castilleja pallida (L.) Spreng. ssp. elegans (Ostenf.) Pennell in Proc. Acad. Nat. Sci. Phil. 86: 526. 1934. Castilleja elegans Ostenf. in Rhodora 36: 187. 1934. Castilleja pallida (L.) Spreng. var. unalaschkensis sensu Ostenf. in Gjda Exp. 64. 1909, non Cham. & Schlecht. See also Porsild in Rhodora 41: 285-286. 1936. Eskimo Lake Basin: Narrows between 3rd and 4th lakes, 3049; Arcric Coast: Cape Dalhousie, 2790; Liverpool Bay, 2929; Great BEAR LakE: Dease Arm, Narakay Isl., 4851; KEEWATIN Distr.: Mistake Bay, 5721. The above series differs from the typical plant in being densely caespitose, often forming large, dense clumps, and in having long yellow wool in the inflorescence. The subspecies elegans is common from the north shore of Alaska to Hudson Bay on the “barren grounds” north of the wooded country. It is not a truly arctic species, however, and seems best developed in the southern parts of its range. Castilleja pallida (L.) Spreng. ssp. typica Pennell in Proc. Acad. Nat. Sci. Phil. 86: 522. 1934. Mackenzie River Devta: Richards Isl., 2119, 2273; Arctic Coast: 6 miles east of Kittigazuit, 2515; Atkinson Pt., 2655; Great Bear Lake: North shore, Haldane R., 49634; McTavish Arm, north shore, 5159, 5184. The above series is rather uniform. The stems are few and ascending from a weak base, puberulent throughout; the leaves are puberulent, narrow, and entire; the inflorescence is short and not very woolly. The ssp. typica is common in sandy places along the Arctic Coast and south to Great Bear Lake, thus covering practically the same range as ssp. elegans. Castilleja Raupii Pennell in Proc. Acad. Nat. Sci. Phil. 86: 528. 1934, fig. on p. 529; Raup in Jour. Arn. Arb. 17: 294. 1936. KEEWATIN Distr.: Yathkyed Lake on Kazan R., 5845. Apparently common from Hudson Bay west to the upper Mackenzie Basin. In the Keewatin District it is a late-flowering southern species. 64 SARGENTIA [4 Castilleja Raupii ssp. ursina Pennell in Proc. Acad. Nat. Sci. Phil. 86: 530. 1934. Mackenzie River Detta: Richards Isl., 21194; Kittigazuit Isl., 2310; East Branch, 7045. The ssp. ursina appears to be more branched than any other northern member of the genus. It seems to be an endemic of the Mackenzie Valley, and the type came from Great Bear River (Miss E. Taylor 86, Can). Castilleja septentrionalis Lindl. in Bot. Reg. 11: tab. 925. 1825; Pennell, Scroph. of E. & Temp. N. Am. 553. 1935. Castilleja pallida var. septentrionalis (Lindl.) Gray, Syn. Fl. N. Am. 21: 297. 1886. Great Bear Lake: Etacho Pt., 1500’ elev., 3513; north shore, near Haldane R., 4963. An eastern species common in temperate and boreal Atlantic North America, west to James Bay and north to southern Baffin Island, with isolated stations west to Alberta. The above stations from Bear Lake constitute the first au- thentic records from the continental Northwest Territories. Euphrasia disjuncta Fern. & Wieg. in Rhodora 17: 190. 1915. Great Bear Laxe: Bear River, 3271, 3371. Erroneously recorded from the Northwest Territories (Mackenzie District) by Fernald & Wiegand, 1. c., based upon Dawson’s collection from Lewes R., in the Yukon Territory (17449, Can). Euphrasia disjuncta is a bicentric species known from Atlantic North America and from the Alaska-Yukon region; the above collection extends the range to the Mackenzie District. Rhinanthus groenlandicus Chab. MACKENZIE River: Near Root R., 62° 30’ N., 6562. Pedicularis arctica R. Br. Chlor. Melv. 32. 1823. For synonymy see Simmons, Vasc. Pl. Ellesmereland 30. 1906, and Hultén, Fl. Kamtch, 4: 116-118. 1930. Since Brown’s name was published there has been much confusion with re- gard to its proper application. Some writers have considered it a synonym of P. Langsdorffii Fisch. ex Stev., while others have considered the plant of arctic Canada a race or variety of the Bering Sea plant. Much of the uncertainty seems to have centered around the question of which name held priority in point of date of publication, and the actual identity of the two plants seems to a large extent to have been neglected. To the writer, who is familiar with both plants in the field, it seems most natural to consider them specifically distinct, even though phylogenetically re- lated. Simmons, I. c., has pointed out that P. arctica “shows great resemblance in habit” to P. hirsuta. As a matter of fact, specimens of P. hirsuta with biden- tate galea are not uncommon in northwestern Greenland and have even been observed in material from arctic Europe (Torne Lappmark, Jukkasjarvi Tesi- vara, 750 m., July 11, 1926, G. Alm, Can). Other characters also, including those of the fruits, show that P. arctica and P. hirsuta are really very closely related, although according to the arrangement used by most monographers of the genus they should be placed in separate sections. It would thus appear to be as logical to consider P. arctica a race or variety of P. hirsuta as P. Langsdor ffi. Pedicularis arctica is a common species along the shores of the Arctic Ocean and of the Arctic Archipelago, reaching the extreme northwest corner of Green- land. Although not a littoral species, it seems to be absent in the interior of the Mackenzie and Keewatin Districts. Pedicularis flammea L. Great Bear Lake: Keith Arm, Deerpass Bay, 3464; Dease Arm, north shore, 4715, 4716, 4903; south shore, Leith Pt. 3590; Keewatin Distr.: Yathkyed L. on Kazan R., 5847; Baker L., 6128. 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 65 An eastern species which in the Northwest Territories reaches to Bear Lake with an isolated station in Alaska. Pedicularis hirsuta L. An eastern species which reaches only to the west shore of Hudson Bay. Hooker (FI. Bor.-Am.) clearly did not understand the genus well, and his record of P. hirsuta from Bear Lake needs confirmation. Pedicularis lapponica L., Macoun & Holm in Rept. Can. Arct. Exp. 5A: 19. 1921. RicuHarpson Mrs.: West of Mackenzie Delta, 6622; MACKENzIE River Deta: East Branch, 6883; Great BEAR LAKE: Dease Arm, Narakay Isl., 4850; north shore, Haldane R., 4964; KeEWATIN Distr.: Yathkyed Lake on Kazan R., 5864; Lake on Tha-anne R., 5601. An eastern species, in our district rare or occasional west to the Mackenzie Delta. Previously known in the Mackenzie District only from a single collection in Coronation Gulf (Can. Arct. Exp.). OROBANCHACEAE Boschniakia rossica (Cham. & Schlecht.) B. Fedtsch. Orobanche glabra Hook. Fl. Bor.-Am, 2: 92. 1838. MackenziE River Detta: Husky R., 6624; East Branch, 7047; Richards Isl., 2120; Eskimo Lake Basrn: 3012; Great Bear Lake: Bear R., Mt. Charles, 3352; McTavish Arm, east shore, 3770, 5319; south shore, 3609, 3635. Common in northern parts of the Mackenzie District in open spruce woods and alder thickets, extending north to the limit of trees. Previously recorded from Great Bear Lake by Hooker, I. c. : LENTIBULARIACEAE Pinguicula villosa L. RicHARDSON Mrts.: West of Mackenzie Delta, 6856; MACKENZIE River DeLta: Richards Isl., 7077; Campbell Lake, 1961; Arctic Coast: 6 miles east of Kittigazuit, 2521; Eskrmo LaKkE Basin: 2nd lake, 3014; Setidgi Lake, 3167; Great Bear Lake: McTavish Arm, east shore, 5358; Keewatin Distr.: Yathkyed Lake on Kazan R., 5841. Rare or occasional in sphagnum bogs of the wooded country north to the limit of trees or slightly beyond. Previously recorded from Great Slave Lake (Raup). Utricularia intermedia Hayne. MACKENZIE River DELTA: Pt. Separation, 1940; Richards Isl., north end, 2275, flowering specimens collected July 22, 1927; Eskimo LAKE Basin: 2nd lake, 2013, flowering specimens collected Aug. 18, 1927; Great BEAR LAKE: Smith Arm, north shore, 50/5; Dease Arm, north shore, 4722; McTavish Arm, north shore, 5282; south shore, 3647; Leith Pt. 3553; KEEWATIN Distr.: Lake on Tha-anne R., 5598. Previously known in our region northward to Great Bear Lake (Richardson). In this district U. intermedia flowers and probably also fruits in favorable seasons. Utricularia minor L. Great Bear Lake: North shore near Haldane R., 5051; McTavish Arm, north shore, 5283; Leith Pt., 3547. Previously known in this district northward to Great Slave Lake (Raup). All specimens were sterile. Utricularia ochroleuca Hartm. MACKENZIE River Detta: South end of Richards Isl., 69° N., 134° 40’ W., with green algae in a wet Carex bog, July 22, 1934, 7076. Utricularia ochroleuca was first recorded from North America west of Green- land by Lily M. Perry, in Rhodora 33: 124. 1931, who collected it on St. Paul 66 SARGENTIA 4 Island, N. S. Our specimens, like those of Dr. Perry, are sterile, but the char- acteristic shape and denticulation of the leaf-segments, the winter buds, and the presence of bladders on the immersed leaf-bearing stems make the determination quite certain. Actually the first collection of U. ochroleuca in Canada came from the north shore of the Gulf of St. Lawrence, Saguenay Co., Quebec, Natashquan, Aug. 7, 1927, H. F. Lewis (Can) (distributed as U. intermedia). Utricularia vulgaris L. var. americana Gray. MAcKENZzIE River Detta: East Branch, sterile but with large winter buds on Aug. 4, 7240, 7289; Esximo LAKE Basin: Setidgi Lake, 3116, 3166; Great Bear Lake: McTavish Arm, north shore, flowering specimens on Aug. 6, 5251. All of these specimens except the last were sterile and the species probably does not flower or fruit except in favorable years. Previously known in the district northward to Great Slave Lake (Raup). PLAN TAGINACEAE Plantago eriopoda Torr. Arctic Coast: Tuktuayaktoq, 7438. As pointed out by Raup (in Jour. Arn. Arb. 17: 297. 1936), Macoun’s refer- ence (Cat. 2: 392) to material from the “Arctic Sea” probably refers to P. sep- tata Morris, and P. eriopoda is thus new to the flora of our region. Plantago juncoides Lam. var. glauca (Hornem.) Fern. in Rhodora 27: 101. 1925. MACKENZIE River Detta: Husky R., June 20, 1931, O. Bryant (Can) ; GREAT BEAR LAKE: McTavish Arm, north shore, by a salt spring, 5187. Apparently new to the flora of our region. Plantago septata Morris in Bull. N. Y. Bot. Gard. 2: 182. 1901. Plantago lanceolata B Hook. Fl. Bor.-Am. 2: 123. 1838; Macoun & Holm, Rept. Can. Arct. Exp. 5A: 20. 1921. MACKENZIE River De_ta: East Branch, 7048; Campbell Lake, 1963, 1964; Kittigazuit Isl., 2396; Arctic Coast: Cape Dalhousie, 2796; Liverpool Bay, 2935; Great Bear LAKE: Bear River, 3394; Dease Arm, north shore, 2723; Dease Valley, 4852. In our district P. septata grows in dry calcareous soil. Our specimens well match the description. Rydberg (Fl. Rocky Mts.) states in his key: “plant not woolly at the base,” but in the text he says “woolly, especially below,” which is correct. Morris, l. c., says: “seeds two.” In our material the capsule is as often 3-locular with one seed in each compartment. Plantago septata appears to be common in Yukon Territory and northwestern Mackenzie. It reaches east to Coronation Gulf and south to the mountains of Alberta and Montana. The species was described from central Yukon Terri- tory, not “Alaska,” as stated in the Gray Index. Unlike P. eriopoda, it is nonhalophytic. RUBIACEAE Galium boreale L. MACKENZIE River DELTA: Campbell Lake, 1962, 2016; Great Bear Lake: Bear R., Mt. Charles, 3338, 3353; Porcupine R., 3372. Galium Brandegei Gray. See Porsild in Rhodora 41: 289, 1939, MACKENZIE River De_ta: Husky R., 7346; Eskimo LAKE Basin: Setidgi L., 3/18; GREAT Bear Lake: McTavish Arm, east shore, 3699, 3709; south shore, 4879; Keewatin Distr.: Lake on Tha-anne R., 5599, 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 67 A common species in wet Carex bogs and on the margins of ponds and sloughs of the lowland, reaching north to the limit of trees or slightly beyond. Not previ- ously recorded from the Northwest Territories. Galium trifidum L. MACKENZIE River Detta: Aklavik, 7313; Eskimo Lake Basin: 2nd lake, 3015; GREAT Berar LaKkeE: Keith Arm, Deerpass Bay, 3444; MAcKENzIE River: Above Sans Sault Rapid, 3401. A southern woodland species in our district, rare or occasional in favorable places north to the tree limit. Previously known northward to Ft. Smith (Raup). CAPRIFOLIACEAE Viburnum edule (Michx.) Raf. MackKENZIE River Detta: Campbell Lake, 2014; Eskimo Lake Basin: Setidgi Lake, 3165; Great Bear Lake: Bear R., 3246; McTavish Arm, 3755, 5252, 5317. Occasional on the lower Mackenzie and at Great Bear Lake. VALERIANACEAE Valeriana capitata Pall. Valeriana bracteosa Britt. in Bull. N. Y. Bot. Gard. 2: 183. 1901. RIcHARDSON Mrts.: West of Mackenzie Delta, 6721, 6858; Arctic Coast oF YUKON TERRITORY: King Pt., 7156; MAcCKENzIE River Detta: Aklavik, 1965; Richards Isl., 2278. An Asiatic species common in Alaska and Yukon, and having its eastern limit at the Mackenzie Delta; Hooker’s statement, “Between the Coppermine and Mackenzie Rivers,” needs confirmation. The plant of the interior is taller and the inflorescence is less “capitate” (V. bracteosa Britt.), but it differs in no other way from lV. capitata. The species is not truly arctic, and north of the limit of trees it does not normally produce mature fruits. Valeriana septentrionalis Rydb. GREAT BEAR LAKE: Etacho Pt., 1500’ elev., 3514. Not previously recorded from the Northwest Territories. CAMPANULACEAE Campanula uniflora L. RicHarpson Mrs.: West of Mackenzie Delta, 6859. A species of peculiar and disrupted range, common in the Eastern Arctic west to Keewatin, in the Arctic Archipelago, and along the Arctic Coast west to Coronation Gulf. Apparently absent in Yukon Territory, with an isolated sta- tion in the Richardson Mts. COMPOSITAE Aster alpinus L. spp. Vierhapperi Onno in Bibi. Bot. 26[ Heft 106]: 25, tab. 6, fig. 4. 1932. Great Bear LAKE: Bear River, dry slopes near summit of Mt. Charles, 3340; Bear Rock, 1300’ elev., 3379; Smith Arm, north shore, 3087. This is A. alpinus of some American authors but not of Linnaeus; it is closely related to A. culminis A. Nels., placed by Rydberg under A. alpinus. From A. alpinus of Europe it differs in having the stem glandular below the head, the basal leaves l-nerved, the bracts more obtuse and purple-tipped, and the rays pure white, not golden yellow or rose. Onno’s description, |. c., is so brief that the plant can hardly be recognized by it, but the plate, as well as specimens annotated by Onno in the Gray Herbarium, leave no doubt as to which plant is meant. 68 SARGENTIA [4 Our plant may be described as follows: Perennial from a branching, stout, sub- ligneous rootstock; basal leaves numerous, oblanceolate, obtuse, 3-6 cm. long, 0.5-0.7 cm. wide, 1-nerved; cauline leaves 5-8, linear, the upper ones much re- duced. Leaves, stems, and bracts invested with very short, scattered strigose hairs, both surfaces of the leaves in addition covered with pale yellowish resinous dots, visible under 25 & magnification. Stems few, stiffly erect, not at all ascend- ing, 10-15 (-25) em. high, dark green with conspicuous reddish ribs; heads solitary, large, 3-4 cm. in diam., 1.5 cm. high; bracts 0.7 cm. long, obtuse, promi- nently tipped and margined with purple; ligules 1-1.4 cm. long and about 0.2 cm. wide, white and papery; achenes flat, cuneate, about 0.3 cm. long, strigose ; pappus sordid, simple. The entire plant in life is viscid-mealy to the touch and the flowering heads are very showy. Distribution: Mountains of Alberta, with isolated stations in Alaska, Yukon, and at Great Bear Lake. Aster Lindleyanus T. & G. Great Bear LAKE: Bear R., Mt. Charles, 3344. Only sterile leaf-rosettes could be found and the species may not flower and fruit here except in favorable years. Previously known northward to Great Slave Lake (Raup). Erigeron acris L. var. asteroides (Andrz.) DC. GREAT Bear LAKE: Keith Arm, Russel Bay, 3413; Dease Arm, north shore, 4726, 4816; McTavish Arm, 3612, 3687, 3756. Erigeron compositus Pursh. Mackenzie River Decta: East Branch, 6895, 7006, 7058; Campbell Lake, 2018; Arctic Coast: Kittigazuit Isl., 2403; Atkinson Pt., 2662; Cape Dalhousie, 2802; Liverpool Bay, 3091; Great BEAR LAKE: Dease Arm, Narakay Isl., 4855; Smith Arm, north shore, 5020; McTavish Arm, north shore, 5318. In addition to the numbers listed above, which are all typical, the following belong to the var. trifidus (Hook.) Gray: Great Bear LAKE: Bear R., Mt. Charles, 3442; McTavish Arm, south shore, 3637, 3659. Erigeron eriocephalus J. Vahl. Arctic Coast: Atkinson Pt., 2663; Cape Dalhousie, 2803; Coppermine R., A. M. Berry 22, 23 (Can); Mackenzie Distr.: Thelon R., 63°-64° N., J. W. Tyrrell 106115 (Can). Our specimens are a good match for typical plants from Greenland. Erigeron grandiflorus Hook. Fl. Bor.-Am. 2: 18, tab. 123. 1834. MACKENZIE River De_ta: Campbell Lake, 1970, 2019; Arctic Coast: Cape Dalhousie, 2804; Liverpool Bay, 2940, 2950. A western species of Alaska and Yukon which reaches eastward along the Arctic Coast to Anderson R., and in the mountains south to high peaks of Alberta. Antennaria angustata Greene, Pitt. 3: 284. 1898; Malte in Rhodora 36: 115. 1934; Polunin, Bot. Can. E. Arct. 347. 1940. Antennaria nitens Greene in Ott. Nat. 25: 42. 1911. KEEWATIN Distr.: Lake on Tha-anne R., 5604, 5605. In its typical form, A. angustata is a well-marked species common in W. Greenland, S. Baffin Island, N. Labrador, and the Hudson Bay region. In Keewatin the species becomes rare and less typical. A form which puzzled both Malte and Polunin is A. nitens Greene, collected in Wager Inlet by J. 1. Macoun (79269, type, Can). It consists of three rather poor individuals collected on Sept. 8, 1910. In the discussion following the description, Greene, 1. c., adds: “quite past flowering . . . and had even shed most of their pappus, enough of which has remained, however, to show their very marked character of being 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 69 almost plumose in the middle and upper portion .. . I can hardly doubt that the plants are all staminate. The appearance of the involucres seem to indicate this, as well as the subplumose pappus.” The three aforementioned specimens, now at any rate, show no single trace of pappus or achenes, and it is evident that Malte, 1. c., when copying Greene’s diagnosis, overlooked the discussion; hence his remark, 1. c., “only the female plant known.” Until better material is available, A. nitens is perhaps best treated as a form of A. angustata, which, through age, has become almost entirely glabrous. In the National Herbarium of Canada are four sheets of a plant from high mountains of Alberta and British Columbia: Alberta: Forget-me-not Mt., J. Macoun 18497 ; Lake Louise, J. M. Macoun 65428 ; British Columbia: Skagit R.., J. M. Macoun 69336, and Columbia R., Spreadborough 19658. These were labelled by Greene with a name which apparently was never published, while they were referred by Rydberg to A. monocephala. All of these, no doubt, should be referred to A. angustata Greene. Antennaria canescens (Lange) Malte in Rhodora 36: 109. 1934. Antennaria alpina (L.) Gaertn. 8 canescens Lange, Fl. Dan. 16: tab. 2786, fig. 1-3. 1869. KEEWATIN Distr.: West coast of Hudson Bay, Mistake Bay, 5723; Yathkyed Lake on Kazan R., 5854; lower Kazan R., 6068. Antennaria canescens appears to be a well-marked species of Greenland, S. Baffin Island, northern Labrador, and the Hudson Bay region, reaching west only to central Keewatin. Antennaria compacta Malte in Rhodora 36: 111. 1934. Antennaria candida sensu Macoun & Holm in Rept. Can. Arct. Exp. 5A: 21. 1921, non Greene. Arctic Coast: Cape Dalhousie, 2801; Great Bear Lake: Bear R., Mt. Charles, 1500’ elev., 3306, 3347; Dease Arm, Narakay Isl., 4853; McTavish Arm, north shore, 5210. The plants from Great Bear Lake are almost pulvinate with short and rather broad, almost imbricate basal leaves; the flowering stems are 10-15 cm. high. The achenes are glabrous. Antennaria crymophila sp. nov. Planta rhizomate brevi percrasso subligneo praedita, rosulas implicatas nume- rosas steriles sessiles edens ; folia basalia spathulata, obtusa, 1 cm. longa, 0.3 cm. lata, per annos persistentia, utrinque conferte adpresse albo-tomentosa, adulta plumbaginea ; rami florigeri ca. 10 cm. alti, rigidi, tomento ablato atro-brunnei, folia caulina 6-8 sat remota valde diminuta apice scariosa gerentes; corymbus laxiflorus, corymbulis 4 pro ratione magnis longe pedunculatis constans; invo- lucrum ca. 0.6 cm. longum, bractearum seriebus 4 vel 5, bracteis basi evidenter atris, apice obtusis, pallide stramineis, pappo subrufo; styli vix exserti; nuculae maturae olivaceae setulis paucis at conspicuis indutis; planta mascula ignota. Mackenzie River Detta: East Branch, 68° 55’ N., dry, gravelly places, July 21, 1934, A, E. Porsild 7053 (tyre). Antennaria crymophila is a well-marked species of the Sect. Dioicoideae (Sched.), perhaps most closely related to A. pulvinata Greene of the mountains of Alberta and British Columbia, but differing from it in the open corymbs, taller habit, and strongly developed rootstocks. Thus far known only from the type locality. Antennaria Ekmaniana nom. nov. Antennaria angustifolia E. Ekman, Sv. Bot. Tidskr. 21: 53, tab. 1, figs. 1, 2, 12. 1927, non Rydb. in Bull. Torr. Bot. Cl. 26: 546. 1899. 70 SARGENTIA [4 Antennaria Friesiana sensu E. Ekman, ibid. 22: 416. 1928, as to plant discussed, not as to type, A. alpina var. Friesiana Trautv. in Acta Hort. Petrop. 6: 24. 1878. Antennaria labradorica sensu Fern. in Rhodora 33: 223. 1931, non Nutt. in Trans. Am. Phil, Soc. new ser. 7: 406. 1841; sensu Malte in Rhodora 36: 112. 1934; sensu Polunin, Bot. Can. E. Arct. 350. 1940, The late Mrs. Elizabeth Ekman, |. c., under A. angustifolia, described and illus- trated a plant which had formerly passed as A. alpina (L.) Gaertn. in Greenland and in North America. Unfortunately the name chosen, which so well described the plant, was preoccupied by A. angustifolia Rydb. Antennaria Friesiana, based upon A. alpina var. Friesiana Trautv., as first suggested by Fernald, 1. c., and later definitely proved by Malte, |. c., was an unfortunate substitution. Fernald further suggested that Ekman’s plant is probably the long obscure A. labradorica Nutt., of which he discovered fragments thus labeled in the British Museum of Natural History, London. In his posthumously published revision of the arctic species of Antennaria, Malte, without contributing any new proof, accepted Fernald’s interpretation of Ekman’s plant. The plant which Nuttall described as A. angustifolia came from the herbarium of Schweinitz, who was bishop of the Moravians, and who undoubtedly had re- ceived it from missionaries in Labrador, whereas Ekman’s plant is truly arctic and, as far as the writer is aware, has not been collected in Labrador proper. Nuttall’s diagnosis is very brief but it contains several statements which seem to show that he could not possibly have had the narrow-leaved, rhizomatous arctic plant in mind. Thus, “sarments procumbent, flowering” generally would be understood to mean long, slender runners, and Nuttall would hardly have used that term if he had wished to describe the sessile innovations so characteristic for the arctic plant. Malte, l. c., 103, who himself stressed the taxonomic impor- tance in this genus of the “presence or absence of stolons,” in the Latin diagnosis of some of his new species translated “basal offshoots” with “sarmentis,” even where sessile innovations were meant. This certainly is very misleading and ‘“iInnovationes” perhaps would have been a more correct term. Nuttall further describes his A. labradorica as having “radical leaves spathulate- linear,” whereas our plant has the narrowest leaves of all arctic species. Of the cauline leaves Nuttall says, “at length nearly smooth,” while in our plants they are permanently lanate, as stated by Malte, 1. c. The identity of our plant with that of Nuttall becomes even more problematical when we read Nuttall’s discussion, in which he says: “Apparently a very distinct species. At first glance resembling A. alpina, but more nearly allied to A. plan- taginea, it is, however, a much smaller plant, the leaves not three-nerved . . .; the achenium is also perfectly smooth; .. .” In view of Nuttall’s comparison of his species with A. plantaginea, Malte’s remark, |. c., 144, loses some of its point when he states that Greene made ‘ta wild guess” when (Pitt. 3: 185), under A. neodioica, he ventured the suggestion that Nuttall with his 4. labra- dorica might have meant that species. As a matter of fact, some specimens of A. neodioica var. gaspensis Fern. in the National Herbarium of Canada match Nuttall’s description of A. labradorica much better than does the narrow-leaved, non-stoloniferous A. angustifolia of Ekman. The same may be said of A. spathu- lata Fern. and A. appendiculata Fern, Malte also claimed that Greene (Pitt. 3: 285) “made another unfortunate guess at the identity of Nuttall’s 4. labradorica” when he compared it with a plant now known as A. ungavensis (Fern.) Malte, collected by Spreadborough in northern Labrador (44442, Can). Since Malte himself points out that the i 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 71 latter plant is stoloniferous and has broad rosette leaves, Greene again stands completely vindicated. : Finally, the achenes in A. labradorica of Fernald and Malte are not “perfectly smooth,” as Nuttall states that those of his plant are; on the contrary the mature achenes in the large number of specimens examined by the writer all show the very characteristic hairs described and illustrated by M. P. Porsild (in Rhodora 33: 218, fig. 4. 1931). These hairs appear to be particularly well-developed in plants from northern Greenland and from arctic northwest Canada. Thus, while the identity of A. labradorica Nutt. is still obscure, there at least can be little doubt that it is not identical with the plant described by Mrs. Ekman as A. angustifolia. For the latter plant the new name A. Ekmaniana is proposed, in recognition of Mrs. Ekman’s very real contribution toward the understanding of this well-marked species. As Mrs. Ekman has already pointed out, this plant is a truly arctic species, perhaps the most arctic in the genus. In Greenland it has not been collected south of Disco Bay. Antennaria isolepis Greene. Great Bear Lake: North shore, 4969; Smith Arm, 5085; Etacho Pt., 1500’ elev., 3516; Keith Arm, Russel Bay, 3440; south shore, Leith Pt., 3558; McTavish Arm, north shore, 5211; Krewatin Distr.: Lake on Tha-anne R., 5603; Yathkyed L., on Kazan R., 5853, 5855. The tomentum of the flowering stems appears to be somewhat viscid and late specimens are often covered with particles of dust and sand. Common or occasional, particularly at some distance from the sea-coast, on sunny slopes where the snow remains late in the spring. General Distribution: From northern Labrador to Alaska, north to Hudson Bay and Great Bear Lake. Antennaria neoalaskana sp. nov. : Planta rhizomate brevissimo ascendente furcato subligneo instructa; innova- tiones sessiles, foliis spathulatis, 1-1.5 cm. longis, 0.2-0.3 cm. latis, apiculatis, utrinque leviter adpresse tomentulosis, demum glabratis; caules florigeri parcius floccosi, 5-8 cm. alti, crassi, saepe subcurvati, haud rigidi, foliis lineari-spathulatis 4—7, summis apice scariosis; corymbi laxi, juventute nutantes, plerumque tri- capitulati, pedunculis brevibus crassis ; involucrum foemineum circa 6 mm, altum, bracteis biserialibus, longe acuminatis, ulterioribus plus vel minus lanosis, juven- tute olivaceo-apiculatis, senectute pallide stramineis; pappi radii circa 4.5 mm. longi, tantum serrulati vel barbellati, nuces setis paucis distinctis munitae; stylo exserto. The staminate plant similar, but heads smaller; involucral bracts ovate- lanceolate, obtuse, pale stramineous; pappus rays somewhat plumose-clavellate. Ricuarpson Mrts.: Eastern slope, west of Mackenzie River Delta, 68° N., 136° W., 4000’ elev., barren gravel ridges near summit, July 7-10, 1933, A. E. Porsild 6727 (TYPE) ; same place, snow-flushes at 2500’ elev., Aug. 16, 17, 1933, 6862, 6864. Perhaps most closely related to A. alaskana Malte, which differs from it by its smaller heads and the very characteristic purplish petioles. Antennaria neo- alaskana superficially resembles A. subcanescens Ostenf., but the shape of the leaves is quite distinct and they also lack the dense, pannose tomentum of that species. Antennaria nitida Greene, Pitt. 3: 283. 1898; Fern. in Rhodora 24: 101. 1926; Raup in Jour. Arn. Arb. 17: 305. 1936. Mackenzie River: Bear Rock, 1300’ elev., 3398; Great Bear Lake: McTavish Arm, east shore, 3739; north shore, 3686; south shore, 3660, 3649. 72 SARGENTIA [4 All specimens collected were of the pistillate plant. New to the flora of the Northwest Territories. Antennaria parvifolia Nutt. in Trans. Am. Phil. Soc. new ser. 7: 406. 1841. Great Bear Lake: McTavish Arm, south shore, 3641; north shore, 5320; east shore, 3738, 3738A. A remarkably uniform series from pre-Cambrian rocks of the east end of the lake is with some doubt referred to this species. Our plants have subligneous rootstocks, rather short innovations, and form small, rather dense mats. The basal leaves are short, less than 1 cm. long, spatulate, somewhat cuneate, densely adpressed-tomentose on both surfaces but not at all shiny as in A. nitida. The cauline leaves are well-developed, 1 to 2 cm. long, and are without scarious appendages. The heads, 3 to 5 in number, are. sessile in an almost globular inflorescence, the outer bracts are lanate below, brown or stramineous, the inner ones obtuse, stiff and papery, white with a tinge of yellow. The style is barely exserted, and the achenes are glabrous and epapillose. Our plants were all pistillate. Antennaria pulcherrima (Hook.) Greene, Pitt. 3: 176, 1897. Antennaria carpathica var. pulcherrima Hook. Fl. Bor.-Am. 1: 329. 1834, and var. humilis Hook. ibid.; Malte in Rhodora 36: 105. 1934, excl. A. lanata; Ostenf. in Gjda Exp. 67. 1910. Antennaria eucosma Fern. in Rhodora 13: 23. 1911, Great Bear Lake: Bear River, 3274; south shore, Leith Pt., 3559, 3598; north shore of Dease Arm, 4724; north shore of McTavish Arm, 5188. Ostenfeld, |. c., from a fragment collected on the Arctic Coast west of the Mac- kenzie Delta, was the first to record this species in the Arctic. The record per- haps needs confirmation, in as much as we failed to collect this rather conspicuous species on the Arctic Coast, in the Mackenzie Delta, or in the mountains west of the Delta. Antennaria pulcherrima appears to be restricted to calcareous soils and for this reason its distribution is not continuous, although it is now known to occur across the continent from the Gulf of St. Lawrence to Alaska, north to Hudson Bay and Great Bear Lake. Notwithstanding Malte’s remarks, 1. c., A. Janata is really amply distinct from A. pulcherrima by its much reduced, conspicuously scarious-tipped cauline leaves, much lower stature, and much more lanate-tomentose leaves and stem. Also, A. lanata is a more southern species with a rather well-defined geographic range. In the large material now available in the National Herbarium of Canada, from mountains of Alberta and British Columbia, there is not a single sheet that in any way may be said to approach A. pulcherrima. Antennaria pygmaea Fern. in Rhodora 16: 129. 1914; ibid. 26: 99, tab. 142, figs. 5 & Sa. 1924; Polunin, Bot. Can. E. Arct. 353. 1940. MacKkENZIE River Devta: Richards Isl., 2286, 2287; Great Bear Lake: Etacho Pt., 1500’ elev., 3515; Krewatin Distr.: Yathkyed Lake on Kazan R., 5852. As correctly pointed out by Polunin, 1. ¢., A. pygmaea is not always mono- cephalous and may even have two smaller heads below the central one (Pt. Burwell, A. E. Porsild 5986). In the writer’s no. 5852 some plants have two heads of equal size side by side. In the interior of the continent 4. pygmaea becomes much taller than in Labrador, and specimens 10-14 cm. high are not uncommon. In all specimens examined the mature achenes are glabrous, the pappus is pure white with a silky sheen, and the styles are barely exserted. 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 73 Antennaria rupicola Fern. in Rhodora 1: 74. 1899. Great Bear Lake: Southern part of McTavish Bay, 3648, 3640; east shore, 37384; Etacho Pt., alpine meadows, 1500’ elev., 3517. This species, originally described from Maine, has since been collected in a number of places about the Gulf of St. Lawrence and in Newfoundland, and lately also from Lake Superior, Thunder Cape, Michigan. Abundant material of a plant first thought to be undescribed was collected in several places on Great Bear Lake. Although closely related to such western species as A. acumunata Greene, A. pedicellata Greene, or A. Howellii Greene, it seems, nevertheless, best placed with A. rupicola Fern., as do some other specimens in the National Her- barium of Canada from Yukon and from northern British Columbia. Although Fernald, ibid. 16: 132. 1914, reduced his species to a variety of A. neodioica Greene, the western plant at least is very distinct from that species. Characteristic for our plant are the strongly developed cauline leaves which are entirely without scarious tips. The achenes are dark brown and distinctly papillose. Antennaria subcanescens Malte in Rhodora 36: 112. 1934; Porsild in Rhodora 41: 295. 1939. Arctic Coast: Cape Dalhousie, 2800. Achillea sibirica Ledeb. Mackenzie River Detta: Portage between Campbell and Setidgi Lakes, 2028. Matricaria ambigua (Ledeb.) Kryl. Fl. Alt. 3: 625. 1904. For complete synonymy see M. P. Porsild in Medd. om Grgnl. 92!: 72. 1932. A common species along the Arctic Coast and thus far not collected in the interior of our district. Chrysanthemum arcticum L. Common along the Arctic Coast and, like the preceding, apparently a littoral species absent in the interior. Tanacetum bipinnatum (L.) Schultz-Bip. MACKENZIE River Detta: East Branch, 7007. Artemisia arctica Less. Artemisia comata Rydb. in N. Am. Flora 34: 263. 1916. Ricuarpson Mrs.: West of Mackenzie Delta, 1500’—2500' elev., 6726, 6869, 6870, 7350; Arctic Coast or YuKoN Territory: Shingle Pt. 6911; King Pt., 7199. This is perhaps the most common Artemisia west of the Mackenzie, on the Arctic Coast and in the mountains. It grows in moist, grassy places, on snow- flushes, and in thickets. Toward maturity the peduncles of the fruiting heads are sharply deflexed. Hooker’s record from Bear Lake, no doubt, refers to A. hyperborea Rydb. Artemisia comata Rydb. is merely the extreme arctic, depressed form of A. arctica Less. The specimens collected by the Canadian Arctic Expedition, in- cluding the Rydberg type from the north shore of Alaska, are all very poor, and, obviously, were collected under snow. Artemisia biennis Willd. MAcKENZIE River: Ramparts above Good Hope, 3404. Artemisia frigida Willd. MACKENZIE RiveR DELTA: Campbell Lake, 2017; East Branch, 6531; Richards Isl., 2122; Arctic Coast: Liverpool Bay, Kugaruk, W. of Anderson R., 3089. 74 SARGENTIA |4 On sandy slopes on the East Branch of the Mackenzie Delta A. frigida is one of the dominant species. When fresh, the plant is strongly aromatic and, during the hot weather, colonies of A. frigida along the East Branch can be de- tected by their fragrance at a distance of several miles. Artemisia hyperborea Rydb. in N. Am. Flora 34: 262. 1916. MACKENZIE River Detta: East Branch, 7052, 7278; Great Bear LAKE: Smith Arm, north shore, 5018; Cape McDonnel, 5164. This species, like A. frigida, when fresh, is strongly aromatic. It grows in dry sandy or gravelly places. Artemisia Richardsoniana Bess. in Bull. Soc. Nat. Mosc. 9: 64. 1836. See Rydb. in N. Am. Flora 34: 261. 1916. Arctic Coast: Cape Dalhousie, 2807. Our specimens are larger but otherwise match ~ : probably a duplicate of the type, labelled by Hooker “A. arctica—Dr. Richardson” (Can): This rare plant, which has been collected but a few times on the Arctic Coast, differs strikingly from A. hyperborea in its dense investment of white wool 0.5 cm. long. The flowering stems are reddish and the inflorescence compact. Ryd- berg’s record, I. c., from “Northern Alberta” probably needs confirmation. Artemisia spithamaea Pursh. Great Bear LAKE: McTavish Bay, 3685. Petasites arcticus sp. nov. Herba perennis, rhizomate crasso horizontali instructa; folia cum floribus contemporanea, palmatim venosa, lobata, reniformia, evoluta 15-30 cm. lata, 10-15 cm. longa, marginibus molliter ciliata caeterum et tenella utrinque glabra, laete pallide viridia, lamina vix ultra medium in lobis 5-8 cuneatis mucronato- dentatis partita, petiolis obscure purpurascentibus quam lamina circa duplo longioribus ; caulis 50-70 cm. longus, parcius floccoso-tomentosus, bracteis scari- osis paucis instructus ; inflorescentia corymbosa, pedunculis fructigeris 5-10 em. longis ; involucrum plantae foemineae 1 cm. longum, bracteis basi tomento tenui brunneo indutis; stigma bilobum, nuculae lineares, glabrae; pappus sericeus albido-flavus. MackENZIE River Detra: East Branch, 68° 40’ N., in wet clay of shady ravines, with almost mature fruit on June 26, 1932, 4. E. Porsild 6515 (rypve). Additional collections are: RicHarpson Mrts.: West of Delta, shaly scree, 1500’ elev., 7349; MAacKENZzIE RIvEeR Detta: East Branch, 6952; Arcric Coast or Yukon Territory: Between King & Kay Pts., 7208; Eskimo Lake Basin: Kugaruk, west of Anderson R., 3058. Petasites arcticus is a common species on wet, shaly, or clayey slopes, where it flowers and fruits very early in the season. It is perhaps most closely related to P. speciosus (Nutt.) Piper, from which it differs in its perfectly glabrous leaves, which also lack the flat bristles on the upper surface of the leaf so charac- teristic of that species. Petasites frigidus (L.) Fries, Macoun & Holm, Rept. Can. Arct. Exp. 5A: 22. 1921. To the synonymy given by Rydb., in N. Am. Flora 34: 312. 1927, should be added: Petasites gracilis Britt. in Bull. N. Y. Bot. Gard. 2: 186. 1901. Petasites frigida (L.) Fries var. corymbosa (R. Br.) Herder, Pl. Radd. Monopet. 32: 4. 1867. Nardosmia sagittata sensu Hook. Fl. Bor.-Am. 1: 307. 1834, non Pursh, at least in part. Arctic Coast or YuKoN Territory: Shingle Pt., 7115; Ricuarpson Mrs.: 6629; Mac- KENZIE River Detta: East Branch, 6961; Campbell Lake, 1971; Richards Isl., 2284, 2285; Kittigazuit Isl., 2405; Arctic Coast: Atkinson Pt., 2669; Liverpool Bay, 2937; Great BEAR Lake: Ft. Franklin, 3225; north shore, Haldane R., 4947. ut 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 7 A western arctic-alpine species in America, common from the Bering Sea region north and east along the Arctic Coast to Coronation Gulf and the western Arctic Archipelago to Melville Island, south to Great Bear Lake and in high mountains of Alaska, Yukon, Alberta, and British Columbia south to lat. 54°. Petasites hyperboreus Rydb. in N. Am. Flora 34: 312. 1927. This apparently is an alpine plant endemic to mountains of Alberta, British Columbia, and possibly Washington. The writer strongly suspects that this species is identical with P. nivalis Greene, Pitt. 2: 18. 1889. Rydberg is cer- tainly wrong when, I. c., he gives Nardosmia corymbosa Hook., Fl. Bor.-Am. 1: 307. 1834, as a synonym, even though he qualifies the statement by adding “mainly.” Although Hooker, |. c., obviously misunderstood that genus, there can be no doubt that by N. corymbosa he referred to the arctic plant. Thus, from the distribution given by Rydberg, |. c., from P. hyperboreus should be excluded : “Arctic Coast from Hudson Bay to Alaska.” Petasites sagittatus (Banks) Gray. To the synonymy given by Rydb., in N. Am. Flora 34: 311. 1927, should be added: Nardosmia frigida sensu Hook. Fl. Bor.-Am. 1: 307. 1834, non L., at least in part. MAackKENzIE River Detta: East Branch, 7297; Aklavik, 7314; Great Bear LAKE: Dease Arm, north shore, 4904; Mackenzie Distr.: Thelon R., lat. 63°-64° N., Tyrrell 23184, 14691 (Can); Thelon R., mouth of Hanbury R., Radford 132367 (Can). It seems abundantly clear that Hooker, |. c., either misunderstood the genus or else his notes somehow had become mixed. Thus Nardosmia frigida is de- scribed: “foliis cordatis inaequaliter dentatis subtus tomentosis,” but in the dis- cussion he states: “From Lake Winnipeg, in lat. 52°, to Ft. Franklin, in lat. 66°, Dr. Richardson—Very few of the specimens gathered have leaves exactly answering to the above character, and of these, some are nearly ten inches long, independent of the petiole.” This clearly shows that Hooker, at least in this in- stance, spoke of P. sagittatus, which is the largest-leaved and least northern of the non-palmately veined species of Petasites. This is amply born out by a photograph in the Gray Herbarium, made by Professor M. L. Fernald in 1903 from Hooker’s Herbarium in the British Museum, showing a sheet of typical P. sagittatus. It is labelled: “Nardosmia frigida Hook. L. Winnipeg R.” An- other photograph shows a sheet which is labelled at the bottom: “Nardosiuia sagittata, Hook. Fl. Bor.-Am., 307 in part, not T. sagittata Pursh.” On the sheet are two collections, both labelled Nardosmia sagittata Hook. The first is from “Cape Bathurst on the Arctic Coast,” the second is labelled “Between Y. H. & Cd. H. Spa. 1825, R.” Both collections show plants of typical P. frigidus in various stages of leaf-denticulation. Finally, a third photograph shows a small specimen of P. sagittatus. It was originally labelled: “Tussilago palmata, Okak in Labrador leg. Weitz”; an additional notation on the back of the sheet reads: “America Septentrionalis, Labrador, 1785, Soc. unit. fratr.” The label bears a subsequent correction which reads: “T. frigida.” Petasites sagittatus is apparently an eastern species distributed from Labrador west to the Mackenzie and Yukon Territory. Arnica alpina (L.) Olin ssp. angustifolia (J. Vahl) Maguire in Madrofio 6: 153. 1942. For his monograph of the genus, Dr. Bassett Maguire has kindly named the Arnicas in the writer’s collection. While the writer would prefer to assign spe- cific rank to ssps. attenuata and tomentosa, he is following Dr. Maguire's interpretation. 76 SARGENTIA [4 Ricuarpson Mts.: West of Mackenzie Delta, 1500’ to 2500’ or occasionally 4000’ elev., 6865; Mackenzie River Detta: East Branch, 6516; Campbell Lake, 1972, 1973; Richards Isl., 2280; Arctic Coast: 6 mi. east of Kittigazuit, 2529, 2530; Tuktuayaktoq, 7447 (the last four numbers, according to Maguire, approach ssp. attenuata) ; Liverpool Bay, 2947; Great Bear Lake: Dease Arm, Narakay Isl., 4856; Bear River, 3248 (last two approaching ssp. attenuata) ; Etacho Pt., 1500’ elev., 3518; Dease Arm, Dease Valley, 4882 (approaching ssp. Sornborgert) ; Dease Arm, north shore, 4728. A widely distributed arctic subspecies, common in North America on sandy and gravelly slopes, from Greenland west at least to the Mackenzie, south to the limit of trees. Arnica alpina (L.) Olin ssp. attenuata (Greene) Maguire in Madrofio 6: 153. 1942. Ricnarpson Mrs.: West of Mackenzie Delta, in thickets on the lower slopes, 7348 (ap- proaching ssp. angustifolia); Mackenzie River Detta: East Branch, 7004; KEEWATIN Distr.: Lake on Tha-anne River, 5602 (approaching ssp. Sornborgert). A western subspecies, perhaps less common and decidedly more southern in its distribution than ssp. angustifolia. Arnica alpina (L.) Olin ssp. tomentosa (J. M. Macoun) Maguire in Madrojfio 6: 153. 1942. Ricnarpson Mrs.: West of Mackenzie Delta, alpine slopes, 4000’ elev., 6728 (X A. lou- iseana ssp. frigida); Arctic Coast: Atkinson Pt., 2670; Great Bear LAKE: Cape Mc- Donnel, 5162A. New to the flora of the Northwest Territories. Arnica Lessingii Greene, Pitt. 4: 167. 1900. Ricwarpson Mrs.: West of Mackenzie Delta, alpine slopes between 1500’ and 2500’ elev., 6866. New to the flora of the Northwest Territories. Arnica louiseana Farr. ssp. frigida (Meyer) Maguire in Madrofo 6: 153. 1942. RicHarpson Mrs.: West of Mackenzie Delta, grassy slopes, 1500’ elev., 6867; Mac- KENZIE River Devta: Richards Isl., 7080; East Branch, 6968; Great Bear LAKE: Cape McDonnel, 5162. New to the flora of the Northwest Territories. Senecio atropurpureus (Ledeb.) Fedtsch. See Porsild in Rhodora 41: 298. 1939. Ricnarpson Mrs.: West of Mackenzie Delta, alpine slopes betweeen 1500’ and 3000’ elev., 6724, 6725, 6875, 6876; Arctic Coast or YuKon Territory: Between King and Kay Pts., 7165, 7166, 7203, 7205. The specimens in the above series are somewhat variable; thus on barren hill tops and in gravelly places the entire plant becomes tomentose, with reduced stem leaves and few heads, while in moist grassy places it is glabrate or sparingly arachnoid with from 2 to 5 heads. In life the rays are dark croceous, but in drying they fade and become much paler. The mature achenes are sparingly strigose-hirsute. A western species in America. Near the Arctic Coast it reaches the west bank of the Mackenzie. Senecio hyperborealis Greenm. in Ann. Mo. Bot. Gard. 3: 98. 1916. Mackenzie River Detta: Campbell Lake, 1967-1969; East Branch, 6518, 7055; Arctic Coast: Liverpool Bay, 2941, 2942; Great Bear Lake: Bear R., Mt. Charles, 3305. This pretty little Senecio grows on calcareous rocks and flowers from the middle of June throughout the summer. It appears to be an endemic of the Yukon Territory and northwestern Mackenzie. Senecio Kjellmanii A. E. Porsild. See Rhodora 41: 299. 1939. RicHarpson Mrs.: West of Mackenzie Delta, 3000’ elev., 6722. “si a i 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES Senecio pauperculus Michx. GREAT Bear LAKE: South shore, Leith Pt., 3557, 3597. Senecio resedifolius Less. RicHarpson Mrs.: West of Mackenzie Delta, 3000’ elev., 6723. Saussurea angustifolia DC. RicHarpson Mrts.: West of Mackenzie Delta, 6877; MacKeNzIE River Detta: East Branch, 6554, 6715; Richards Isl., 2131, 2293, 7452; Kittigazuit Isl., 2406; Eskimo LaKE Basin: 3016; Setidgi Lake, 3171, 3172; Great Bear Lake: Dease Arm, Dease Valley, 4881; Krewatin Distr.: Yathkyed L., on Kazan R., 5859. The above series shows a great deal of variation from glabrous to lanulose and from narrow, entire to lanceolate, repand-dentate leaves. Some are low with a few heads in open, long-pedunculate corymbs, others are from 8 to 50 cm. high. Since they are all alike in floral characters, the writer thinks it inadvisable to separate the material on vegetative characters alone. A western arctic species, in America reaching east to central Keewatin. Taraxacum The large series of Taraxacum collected in Alaska in 1926 and in the Mac- kenzie District in 1927-1928 were sent to the late Dr. Dahlstedt in 1930, but upon his death they were returned unnamed. In 1936, the collection, with addi- tional material collected in the Keewatin District in 1930 and in the Mackenzie in 1932-1935, was sent to Dr. Haglund. When returned two years ago, anno- tations showed that in the collection Dr. Haglund had detected 5 or 6 undescribed species. As far as the writer is aware, none of them has yet been published. Taraxacum alaskanum Rydb. RicHarpson Mrs.: West of Mackenzie Delta, 1000’—4000’ elev., 6878, 7347; MACKENZIE River Detta: East Branch, 6557; Richards Isl., 2295. A very pretty little species, apparently endemic to Alaska and Yukon, and reaching east to the Mackenzie or slightly beyond. Taraxacum Carthamopsis M. P. Porsild in Trans. Roy. Soc. Can. 3 ser. 33: 29, tab. 1-3, 1939. Mackenzie River Detta: East Branch, sandy river banks, 7279. Thus far known only from the type locality. Taraxacum lacerum Greene. GREAT Bear Lake: Dease Arm, Narakay Isl., 4857; McTavish Arm, north shore, 5322. Taraxacum phymatocarpum Dahlst. Arctic Coast: Cape Dalhousie, 2812; Liverpool Bay, 2951. Rare or occasional on the Arctic Coast and Islands. Agoseris ? cuspidata (Pursh) Steud. MackENZIE River De_ta: East Branch, between 68° 30’ and 68° 55’ N., 6741, 6885, 6896, 7051. Our plant has narrow, linear-lanceolate leaves, 0.8 cm. wide and 20-25 cm. long, shorter than the scape, pubescent and conspicuously ciliate, purple at the base ; scape lanulose below the head; bracts broad, cuspidate with dark purplish center, densely and coarsely lanulose. The ligules are yellow, fading toward the tip. Achenes cylindrical, beakless, prominently ribbed. Our plant flowered in the last week of July and produced abundant fruit. It is rare or occasional on sandy slopes along the East Branch and was seen nowhere else. 78 . SARGENTIA 4 Crepis elegans Hook. Great Bear Lake: Bear R., 3249, 3373. Crepis nana Richards. RicHarpson Mrs.: West of Mackenzie Delta, 6627. Hieracium gracile Hook. Great Bear LAKE: Etacho Pt., 1500’ elev., 3524. New to the flora of the Northwest Territories. Hieracium canadense Michx. Great Bear Lake: Bear R., Mt. Charles, 3345; McTavish Arm, 3658. BIBLIOGRAPHY Bell, R. 1884. Observations on the geology, mineralogy, zoology, and botany of the Labrador coast, Hudson’s Strait and Bay. Geol. Surv. Can. Rept. Prog. 1882- 1884, DD. Brown, Robert. 1824. Chloris Melvilliana, Suppl. App. No. XI, Capt. Parry’s Voy. 1819-20. Clarke, C. H. D. 1940. A biological investigation of the Thelon Game Sanctuary. Nat. Mus. Can. Bull. 96, Biol. Ser. 25. Fernald, M. L. 1925. Persistence of plants in unglaciated areas of boreal America. (Mem. Gray Herb. 2); Mem. Am. Acad. 15 (3). Grgntved, Johs. 1936. Vascular plants from arctic North America, etc. Rept. Fifth Thule Exp. 1921-1924, 2 (1). Holm, Theo. 1922. Contributions to the morphology, synonymy, and geographical distribution of arctic plants. Rept. Can. Arct. Exp. 1913-1918, 5 (B). Hooker, W. J. 1829-1840. Flora Boreali-Americana. Hultén, E. 1937a. Flora of the Aleutian Islands. —. 1937b. Outline of the history of arctic and boreal biota during the Quarternary Period. Johansen, Frits. 1924. General observations on the vegetation. Rept. Can. Arct. Exp. 1913-1918, 5 (C). Macoun, J. M. 1911. Flora and fauna of west coast of Hudson Bay. Summ. Rept. Geol. Surv. Can. 1909-1910, 281-283. —— and Theo. Holm. 1921. Vascular plants. Rept. Can. Arct. Exp. 1913-1918, 5 (A). Ostenfeld, C. H. 1909. Vascular plants collected in Arctic North America by the Gjoa Expedition under Captain Roald Amundsen, 1904-1906. Vidensk.-Selsk. Skr. I. Math.-Naturv. KI. 8. Polunin, N. 1940. Botany of the Canadian Eastern Arctic, Pt. I. Pteridophyta and Spermatophyta. Nat. Mus. Can. Bull. 92, Biol. Ser. 24. Porsild, A. E. 1929. Reindeer grazing in northwest Canada. Dept. Int. Can., N. W. T. and Yukon Branch. ——. 1937. Flora of the Northwest Territories. In Canada’s Western Northland, 130-141. Can. Dept. Mines & Res., Lands, Parks & Forests Branch. —. 1939. Contributions to the flora of Alaska. Rhodora 41: 141-183, 199-254, 262-301. Preble, E. A. 1902. A biological investigation of the Hudson Bay region. U. S. Dept. Agr. Biol. Surv. N. Am. Fauna No. 22. —. 1908. A biological investigation of the Athabaska-Mackenzie region. U. S. Dept. Agr. Biol. Surv. N. Am. Fauna No. 27. Raup, H. M. 1935. Botanical investigations in Wood Buffalo Park. Nat. Mus. Can. Bull. 74, Biol. Ser. 20. ——. 1936. Phytogeographic studies in the Athabaska—Great Slave Lake region, I. Catalogue of the vascular plants. Jour. Arn. Arb. 17: 180-315. Richardson, J. 1823. Botanical appendix. Franklin’s narrative of a journey to the shores of the polar sea, etc. 1943] PORSILD, FLORA OF THE NORTHWEST TERRITORIES 79 —-,. 1851. Arctic searching expedition: a journal of a boat voyage through Rupert’s Land and the arctic sea in search of the discovery ships under command of Sir John Franklin. Simmons, H. G. 1906. The vascular plants in the flora of Ellesmereland. Rept. 2nd Norwegian Arct. Exp. in the “Fram,” 1898-1902, No. 2. ——. 1913. A survey of the phytogeography of the arctic American archipelago. Lunds Univ. Arssk. N. F. Afd. 2, 9 (19). Tyrrell, J. B. 1897. Report on the Doobaunt, Kazan and Ferguson Rivers and the north-west coast of Hudson Bay, etc. Geol. Surv. Can. 9 (F). Tyrrell, J. W. 1898. Across the subarctics of Canada. NationNAL HERBARIUM OF CANADA, OTTAWA. Saat 2 | ed a 7 : ai a 1 - 4 "v4 4 . on | ~ L. ' : ~ 7 a . . - - : ’ ~_ 2 : ; : ; : : = : : a - oe . A 7 7 5 : . . | J . | F s ; ; i av. 7 ; ; ’ ad - 7 : ~ | a . , : ; : « ¥ : e eee: ; a : ; 7 } as cal 7 ; . « mee | } a 7 « a0 . ’ ” : . | 7 s 7 5 - - — —— 7 — a a se 7 - » ans oe |} hal 7 _ - - 7 é 7 - : - = i = — eo) = = bes : 7 ~ - - - ; ; A 5 : ¥ i fi «1 , A : ia Ny ¢ : - LA oa 7 7 = 7 a ma © M » _ a a : P ; | = * : i i - : = ; : . : ‘24 - - . : 5 ” _ 7 Ce sr ? as ee 7 4 7 7 ‘golf ee = : 7 = : ey Ye 3 a —— ’ sa 7 = oe = a L 2 7 a a ¢ — —— fant > a ’ ' : THE WILLOWS OF THE HUDSON BAY REGION AND THE LABRADOR PENINSULA Hucu M. Raup with four plates INTRODUCTION Tue following paper, on the willows of the Hudson Bay region and the Labrador Peninsula, has grown out of efforts to determine collections from that area. In the summer of 1939 Dr. and Mrs. Ernst C. Abbe and Mr. John Marr, of the University of Minnesota, made a journey to Richmond Gulf on the eastern coast of Hudson Bay for the purpose of making general botanical col- lections.'. Later, in working up his material, Dr. Abbe obtained a quantity of undetermined or partially determined material from the Hudson Bay region collected by others over a period of many years. Most of this came from the Herbarium of the Carnegie Museum of Pittsburgh and was collected by various people who have gone to Labrador and Hudson Bay from that museum on ornithological and mammalogical collecting expeditions. Most of the botanical specimens were gathered by Mr. W. E. Clyde Todd, Dr. George M. Sutton,? and by Mr. and Mrs. J. K. Doutt.* In addition to the Carnegie Museum material Dr. Abbe also obtained some Labrador collections from the University of Minne- sota, mostly collected by Margaret E. Oldenburg, who made a trip around the Labrador coast in 1939, The willows from these assembled collections, totalling 284 numbers, were turned over to me for determination. With the exception of one collection from Southampton Island, scarcely any records for them have previously appeared in the literature, and they constitute most of the new material on geographic dis- tribution to be found in the present paper. The exception mentioned above is that of the collection of Dr. George M. Sutton in 1929-30, of which I published a list in 1936.4 This collection was restudied later by Dr. Nicholas Polunin and cited in his “Botany of the Canadian Eastern Arctic.” ® In working up this material and preparing the list I have had access to the collections of the Gray Herbarium, the Arnold Arboretum, and the National Herbarium of Canada. I have also seen much of the pertinent material at the herbarium of the New York Botanical Garden, particularly of those species which are of critical nature. In all I have seen about 1300 collection numbers, involving approximately 1700 specimens. 1 Abbe, E. C. The expedition to Hudson Bay of the University of Minnesota. Science 90: 458-459 (1939). 2 See Mem. Carn. Mus. 12. ’ See Todd, W. E. C. More about Labrador. An account of a trip to the Grand Falls of the Hamilton River. Carnegie Mag. 13: Sept. (1939). Also, Ann. Rept. Carn. Mus. Pittsburgh 1938: 14-15 (1939). 4 Raup, H. M. Pteridophyta and Spermatophyta of Southampton Island. Mem. Carn. Mus. 123: 17-30 (1936). 5 Polunin, Nicholas. Botany of the Canadian Eastern Arctic, I. Pteridophyta and Sper- matophyta. Nat. Mus. Can. Bull. 92: Salix, 151-173 (1940). See also Polunin, N. Flora of Southampton Island, Hudson Bay. Jour. Bot. 76: 93-103 (1938). 82 SARGENTIA |4 The area involved covers the shores of Hudson and James Bays and the Labrador Peninsula. Westward in the District of Keewatin it includes the basins of the Kazan River and Baker Lake and the basin of the Dubawnt River up as far as Dubawnt Lake. In northern Manitoba and Ontario it extends westward approximately to long. 96° W. and southward to about lat. 51°. In the north it includes Southampton Island and smaller islands of the northern part of Hudson Bay, as well as the islands in Hudson Strait (in the District of Franklin). Southeastward I have included all of the Labrador coast south to the Straits of Belle Isle, and the north shore of the Gulf of St. Lawrence westward to Seven Islands. The islands off the north shore of the Gulf, including Anti- costi, have also been covered, as well as that part of Newfoundland which adjoins the Straits of Belle Isle. It will be noted that this area adjoins or overlaps the areas of the following comprehensive lists recently published: at the north, Polunin’s “Botany of the Canadian Eastern Arctic”; at the southwest, Rydberg’s “Flora of the Prairies and Plains’;® and at the southeast, Marie-Victorin’s ‘Flore Laurentienne.” * Polunin’s area extends south to lat. 60° and west to long. 95°, thus forming the largest overlap with the present paper. All of the willows listed by Polunin for the eastern Arctic, except for a minor form described from southern Baffin, are to be found in the area treated here. Rydberg’s Flora reaches northward, tech- nically, only to southern Manitoba in this longitude, but so many of the willows listed in it are of wide range in the Canadian forests that it serves very well for much of the country in the whole basin of Lake Winnipeg. Victorin’s flora covers southern Quebec northeastward only to Lake St. John and the Saguenay River. Eastern Keewatin will be more thoroughly treated when A. E. Porsild’s studies of the western arctic Canadian flora are published and his specimens distributed. This work is now in progress. The present paper contains very little in the way of taxonomic innovation, and no new species, varieties or forms. In general organization it rests very largely upon the monographic studies of American willows made by Dr. Camillo Schneider. Its main reason for being is that it records for the first time in one place a large mass of new fact about the willows of this region. The imme- ciate results have been the clarification of some of the more poorly known species and a greatly increased knowledge of the geographic aspects of the genus as a whole in northeastern America. It is hoped that these results, together with the diagnostic keys, will facilitate further collection and study of the group. The accumulation of large amounts of new specimen material, however, is not an unmixed advantage; for while some of the lesser-known entities are clarified, many of the “well-known” ones begin to fray at the edges and to show distress- ingly close relationships with other ‘‘well-known” species. This seems to be particularly the case with the genus Salix, although whether it is due, as some students appear to think, to promiscuous hybridity, or to some other form of cytological perversity, is yet to be fathomed. I shall make no attempt to give a complete history of willow-collecting in the region covered. In the middle of the last century the great Swedish mono- grapher, N. J. Andersson, had only a few scattered specimens from the southern Labrador coast and a few from the arctic and Hudson Bay coasts. The first 6 Rydberg, P. A. Flora of the prairies and plains of central North America. Publ. by the N. Y. Bot. Gard. (1932). 7 Victorin, Fr. Marie-. Flore Laurentienne. Montreal (1935). 1943] RAUP, WILLOWS OF HUDSON BAY REGION AND LABRADOR 83 modern collections of consequence were those of the Canadian botanists and geologists John Macoun, James M. Macoun, William Spreadborough, Robert Bell, and A. P. Low. The knowledge of eastern arctic and subarctic willows accumulated by M. S. Bebb in the 1880’s and ’90’s, and by Rydberg in 1899, rested largely upon these Canadian collections. They were of necessity scattered and incomplete. Spreadborough and J. M. Macoun made trips with geological survey parties to James Bay and the southwestern coasts of Hudson Bay, partly by boat, partly overland to the southern tip of James Bay and northeastward from Lake Winnipeg. James M. Macoun did not make his collecting trip along the northwest coasts of Hudson Bay until later, in 1910.5 The coasts of the Labrador Peninsula were surveyed by Bell and Low and their associates in the 1880's and 1890's, for the Canadian Geological Survey. Both these men were remarkably well-trained naturalists, and I have seen nearly 50 numbers of willows collected by them alone. On Low’s trip from Richmond Gulf to Ungava Bay in 1896 he took along Spreadborough, who made substantial botanical collections from a vast inland region not since visited by a botanist. Dr. Camillo Schneider, a German student of willows, came to America in 1917 and was interned at Boston for the duration of the first World War. He was quartered at the Arnold Arboretum and undertook monographic studies of American Salix, which he finally finished in 1921, after his return to Germany. While at the Arboretum he borrowed large quantities of material from the prin- cipal herbaria of America, making careful dissections and sketches of hundreds of specimens. At the same time, although greatly handicapped due to the in- accessibility of types, he made remarkable progress toward the solution of many problems of nomenclature. Schneider brought to bear, in these studies, an extraordinary knowledge of the genus as a whole. Just before beginning the American work he had finished a synopsis of the genus in eastern Asia and the Himalayas for “Plantae Wilsonianae.”*® In America he was able to see repre- sentative material from nearly all parts of the continent from which specimens were then available. His delineation of American species, although far from ideal in many instances, is in the main realistic and workable; and his compre- hensive keys are the best that have been published to date. Unfortunately, his results were printed serially !° in such a way that they have reached relatively few students, and under a reprint policy that has made it almost impossible to gather the whole series into a single book. The first few volumes of the Journal of the Arnold Arboretum have long since been out of print, so that Schneider’s keys, index, and much of his discussion are entirely unobtainable. Furthermore, the keys are all in Latin, of which translations have never been published, so that students and collectors who do not use Latin readily are handicapped at the outset. With regard to our region, however, Schneider was not much better supplied with material in 1918 than were Bebb and Rydberg in the ’90’s. Of the 1300 8 Macoun, J. M. Flora and fauna of the west coast of Hudson Bay. Geol. Surv. Can. Summ. Rept. 1910: 281-283 (1911). 9 Schneider, C. Salicaceae, in Sargent, Plantae Wilsonianae 3: 16-179 (1916), Publ. Arn. Arb. 4. 10 Schneider, C. A conspectus of Mexican, West Indian, Central and South American species and varieties of Salix. Bot. Gaz. 65: 1-41 (1918); Notes on American willows, I. Bot. Gaz. 66: 117-142 (1918), II. ibid., 66: 318-353 (1918), III. ibid., 67: 27-64 (1919), IV. ibid., 67: 309-346 (1919), V. Jour. Arn. Arb, 1: 1-32 (1919), VI. ibid., 1: 67-97 (1919), VII. ibid., 1: 147-171 (1920), VIII. ibid., 1: 211-232 (1920), IX. ibid., 2: 1-25 (1920), X. ibid., 2: 65-90 (1920), XI. ibid., 2: 185-204 (1921), XII. ibid., 3: 61-125 (1921). 84 SARGENTIA |4 numbers that I have seen, only 234 were available to Schneider in 1917-18, and probably not more than 50 of these were new after Rydberg’s study in 1899. In short, during the past 25 years, since Schneider’s monographic work, the willow collections from our region have been increased more than five-fold. To the greater numbers of specimens, moreover, must be added the fact that far more highly trained and experienced collectors have visited the region in these recent years than in all the preceding period. To mention only a few, there were M. O. Malte, E. C. Abbe, R. H. Woodworth, M. L. Fernald (Nfld.), Nicholas Polunin, Victorin & Rolland, H. F. Lewis, and Harold St. John. During the past 25 years, also, several new species and varieties have been de- scribed, chiefly by Professor Fernald, who has done much to straighten out some of the worst taxonomic tangles with which we have to deal. It has been necessary to re-examine these recently described plants and insert references to them in the keys. Neither in the arrangement of the list nor in the keys have I indicated sectional subdivisions of the genus. With our present understanding of relationship, present and past, among the species, it is somewhat hazardous to do so. True, some of the species fall readily and reasonably into groups that can be regarded as sections, but intermediates can be recognized between so many of these sup- posedly well-established sections that their boundaries in many cases cease to be useful. Furthermore, the continued description of new forms not only in- creases the number of these intermediates, but also continues to bring out species that will not go into any of the known sections. For what it may be worth, the following is a tentative disposition of the species of our region in the recognized sections of the genus. PENTANDRAE: serissima, lucida. LONGIFOLIAE: interior. RETICULATAE: reticulata, vestita, jejuna, HeErBACEAE: Uva-ursi, herbacea. OVALIFOLIAE: arctica, arctophila, hudsonensis. GLAUCAE: glauca, cordifolia, brachycarpa, fullertonensis. CorpDATAE: cordata, lutea, myrtillifolia, glaucophylloides, adenophylla. BALSAMIFERAE: pyrtfolia. CHRYSANTHEAE: Richardsonii, calcicola, Wiegandii, alaxensis, laurentiana. CANDIDAE: candida, cryptodonta. FutvaE: Bebbiana. RosEAE: pedicellaris, simulans. GRISEAE: humilis. DiscoLores : discolor. PHYLICIFOLIAE: planifolia, paraleuca, pedunculata, amoena, pellita. Sect. INCERTAE: arbusculoides, argyrocarpa. Some of the uncertainties of this arrangement are shown by S. simulans, which combines the leaves of the Roseae with the pistillate aments of the Glaucae; by S. jejuna, which has characters of both the Reticulatae and the Ovalifoliae; by S. amoena, which has a general resemblance to the Phylicifoliac, but with floral bracts like those of the Glaucae. As pointed out by Schneider, S. arbusculoides and S. argyrocarpa are more or less unique and do not fit into any known sections. The maps " are based principally upon specimens cited in this paper, although in some cases published records have been used if they could be reasonably well authenticated. Very few such records have been used for the more variable and 11 The base map used is by J. Paul Goode, published by the University of Chicago Press. 1943] RAUP, WILLOWS OF HUDSON BAY REGION AND LABRADOR 85 controversial species groups. I have tried to make the maps as complete as possible for the whole area of each. For regions outside that treated in the present paper I have used, in addition to specimens at the Harvard herbaria, chiefly Polunin’s “Botany of the Canadian Eastern Arctic,” Grgntved’s paper on collections from the 5th Thule Expedition,'* and papers by M. P. Porsild ** and T. W. Bocher '* on the Greenland flora. A detailed discussion of problems in the geographic distribution of north- eastern American willows is beyond the scope of the present paper. Neverthe- less, a number of outstanding geographic facts are evident even from a cursory examination of the maps, and may be stated briefly. Sixteen of the species listed are distinctly southern in their distribution within our region, reaching northward only to the north shore of the Gulf of St. Lawrence or at most to the southern parts of James and Hudson Bays or the Hamilton River district. Some of these are of wide western range in the Canadian forest: interior, brachycarpa and var. antimima, pyrifolia, myrtillifolia, pedicellaris var. hypoglauca, Bebbiana, candida. Others are eastern in_ the Canadian forest or range far southward in the Alleghanian region: serissima, lucida, cordata, glaucophylloides, adcnophylla, humilis, discolor, pellita. Salix lutea probably should be in this category, although it apparently comes into our region from the northern Great Plains, where it has a large development. Some of the more distinctly eastern forms have close relatives, or vicariads, in western America. The eastern serissima and lucida are replaced in the west by lasiandra ; cordata by mackensicana, pseudocordata, and their relatives; discolor by Scou- leriana; pellita by subcoerulea; Bebbiana by Geyertana. At the other extreme, in the north, is a group of arctic species, some of which come south only to the Straits of Belle Isle or Newfoundland while others, such as herbacea and Uva-ursi, reach the mountains of New England. Only one of these is of wide range in arctic America: reticulata; the others are prevailingly eastern: arctophila, arctica vars. Browneit and kophophylla, herbacea, Uva-ursi, argyrocarpa, vestita (typical), cordifolia and vars., calcicola. As in the case of the southern species, some of these have western counterparts, arctica vars. Brownei and kophophylla being replaced in the west by other forms of arctica, typical vestita by its var. erecta, cordifolia by glauca and its variations, and calcicola by Richardsonii. Some of the western vicariads reach eastward to Hudson Bay and overlap their eastern relatives in that region, 1.e., glauca, Rich- ardsonii, and vestita var. erecta. Another willow which is widespread in the eastern arctic is S. planifolia, which, however, extends far west in the northern Canadian forest to the Rocky and Mackenzie Mountains. It is finally replaced in Mackenzie, Yukon, and Alaska by S. pulchra, and farther south in the moun- tains by its var. monica. In the northwestern and western parts of our area there appears a peculiar group of species, some of which seem to be endemic in central arctic America while others are of far western affinity. Of the latter, Richardsonti and glauca have already been mentioned, while arbusculoides and alaxensis var. obovalifolia 12 Grgntved, Johs. Vascular plants from arctic North America collected by the Fifth Thule Expedition, 1921-24. Rept. 5th Thule Exped. 2: 32-35, Copenhagen (1936). 18 Porsild, M. P. Stray contributions to the flora of Greenland. VI-XII. Meddel. Grgnl. 933: 60 (1935). 14 Bocher, T. W. Biological distributional types in the flora of Greenland. Meddel. Grénl. 1062: 1-339 (1938). 86 SARGENTIA [4 should be added. None of these have ever been collected on the Labrador coast or the north shore of the Gulf of St. Lawrence, although Richardsonii and glauca var. acutifolia have been found in southern Baffin, and the latter on the shores of Hudson Strait. Apparently endemic to this region and the central arctic coast are fullertonensis, hudsonensis, calcicola var. Nicholsiana, and glauca var. steno- lepis. It may be that alaxensis var. obovalifolia should be in this small group of endemics. Ball, who described the variety, believes that it is peculiar to the country east of the northern Cordillera. It should be noted that all of the endemic group are rather poorly known and need more study before their taxo- nonuc and geographic affinities can be properly worked out. There remains to be mentioned a group of apparent endemics and _ isolated populations that center about the Gulf of St. Lawrence.'* The endemics include jejuna, Wiegandii, laurentiana, cryptodonta, simulans, paraleuca, pedunculata, amoena, vestita var. psilophylla, cordifolia var. eucycla, myrtillifolia var. brachy- poda, pellita f. psila. The presence of typical S. brachycarpa in Gaspé, Anticosti, and on Hudson Bay is a striking example of a Rocky Mountain species with isolated populations in the east. It is of interest to consider, together, the group of arctic and subarctic species which are distinctly northeastern in America. In addition to the 12 endemics of the Gulf of St. Lawrence just noted, there are Uva-ursi, argyrocarpa, and cordifolia that have not been authentically recorded west of the east coast of Hudson Bay. Another small group has been collected on the west coast but not much farther west: arctica vars. Brownei and kophophylla, herbacea, vestita (typical), calcicola. To the above must be added the two endemics of the northern part of Hudson Bay and southern Baffin, calcicola var. Nicholsiana and glauca var. stenolepis. Salix fullertonensis and hudsonensis may have to be included here, but until their western ranges are better known they can hardly be considered with the eastern group. One can hardly avoid an immediate division of these 22 eastern forms into two categories by the size of their areas. The endemics of the St. Lawrence and the Hudson Strait region are all highly localized, with areas relatively very small. All of the other species noted, on the other hand, are of wide range or are nearly ubiquitous. Another fact of interest is that the relationships of the Gulf of St. Lawrence endemics are not with the group of species that has a wide southern range in the Alleghanian region, but rather with arctic and Canadian forest species. Salix jejuna has its nearest relations with reticulata; Wiegandii and laurentiana probably belong in the section Chrysantheae, which contains alaxensis, Richardsonii and calcicola; cryptodonta was placed provisionally in the Candidae by Schneider; paraleuca, pedunculata, amoena, and pellita apparently belong in the Phylicifoliae with planifolia, pulchra, ete. The keys are largely modifications of those already in use. The main key to the species follows Schneider’s,’® with the insertion of forms that have been described since he prepared it. Rather than confuse the main key with varietal subdivisions, I have placed keys to varieties and forms under the separate species. Some of these minor keys are newly constructed or modified from existing ones, while others, such as that for the varieties of S. cordifolia, I have copied verbatim from their sources. 15 See Fernald, M. L. Persistence of plants in unglaciated areas of boreal America. (Mem. Gray Herb. 2) ; Mem. Amer. Acad. 15° (1925). 16 Schneider, C. Jour. Arn. Arb. 3: 97-106 (1921). 1943] RAUP, WILLOWS OF HUDSON BAY REGION AND LABRADOR 87 A key involving staminate flowers would be extremely desirable, but it could be only partially practicable with our present knowledge of the willows of this region. This is due, first, to inherent difficulties in the construction of such a key, for the degree of variability among the characters of the staminate flowers is not nearly so well-known as among those of the pistillate. Even with a rela- tively large amount of material before him, Schneider had only indifferent success with his key to staminate plants and found it nearly impossible to deter- mine, by means of staminate material, many of the species in such a widespread and variable group as that related to S. glauca. A second major difficulty arises from the fact that, to the best of my knowledge, fully one-fourth of the species treated in this paper are known only by their pistillate plants: cryptodonta, lauren- tiana, paraleuca, pedunculata, amoena, pellita, simulans, jejuna, hudsonensis, fullertonensis. True, most of these are among the rarer species, but S. pellita is widespread and well-represented in the herbaria. Some of the species flower very early in the season, so that the ephemeral staminate blossoms are gone before the dates when collectors usually reach the field. Any workable key to the willows perhaps needs some defense. An ideal diag- nostic key to an ideal genus uses clear-cut unit characters that immediately sep- arate the species. However, so many species of willows are based upon elaborate combinations of characters that such an ideal situation is scarcely ever realized. Schneider fully appreciated this and constructed a key that is to a great extent descriptive ; consequently it does not depend so heavily upon individual characters. Those who have not used it extensively, or who have not had to deal with large series of specimens, are apt to be impatient with its length and complexity, not to mention the extraordinary amount of apparent equivocation that occurs in it. In a sense, these words and expressions of equivocation are an index to the practicality of thé key, because they allow for a large amount of character varia- bility while the composite pattern of the species is being outlined. I have made no effort to present the complete synonymy of each species. Following the name are cited the place of original description of the species, variety or form, and then the papers in which this name has been used in describing the flora of our region. These papers are cited in chronological order. Likewise, such synonyms as I have mentioned are those that have been used in floras of the region. For more complete synonymy the student is referred particularly to the cited papers by Schneider, and to those by Dr. Carleton R. Ball for the sections Cordatae, Longifoliae, and Pentandrae. In the citation of specimens, I have followed a geographic sequence beginning in the northern part of Hudson Bay with Southampton Island. Then come the western coasts of Hudson and James Bays southward to the tip of the latter, and then the eastern coasts northward. These are followed by records from the south side of Hudson Strait, Ungava Bay, and the coast of Labrador southward to the Straits of Belle Isle. The sequence then proceeds westward along the north shore of the Gulf of St. Lawrence to Seven Islands, after which come the Mingan Archipelago and Anticosti. Specimens from inland regions along this course, or islands off the coasts, are usually cited in their appropriate places ; thus plants from inland Keewatin are cited opposite those from the coast in that region, and those from the Hamilton River country are usually cited opposite those from Hamilton Inlet and vicinity. Plants from northern Newfoundland are grouped with those on the Labrador side of the Straits of Belle Isle. I have 88 SARGENTIA |4 also inserted the names of territories and provinces, dividing the Quebec localities rather arbitrarily into “Northern’—those on James and Hudson Bays and Hudson Strait, and “Southern’”—those on the north shore of the Gulf of St. Lawrence. In the case of islands, their correlation with political boundaries may not be infallible, for I have preferred to cite the specimens from these islands along with that part of the coast which they are nearest. I have been able to locate on maps nearly all of the localities from which I have seen speci- mens, but there are still several (less than half a dozen) that I have not found. These are cited only approximately in their proper places in the geographic sequence, In each citation the name of the locality is given in roman type, and the name of the collector, with the number or numbers, in italics. If no number is avail- able, dates have been used. The herbaria from which I have seen the specimens are indicated by symbols in parentheses. If more than one number of the same plant, locality, and collector were seen, these symbols appear only after the last number. A key to the symbols is as follows: A—Arnold Arboretum: C— Carnegie Museum; G—Gray Herbarium; M—University of Minnesota; N— New York Botanical Garden; O—National Museum of Canada. No claim whatever is made to finality in the naming of critical specimens. There are so many of these intermediate forms, and their positions in relation to the principal species complexes are often so problematical, that at best the disposition I have made of them can represent little more than a series of opinions. Someone with a slightly different concept of the major species would arrive at entirely different conclusions. I am deeply indebted to a number of persons and institutions for assistance. Thanks are due particularly to Dr. Abbe for entrusting me with his collections and for his patience with delays in working them up. Professor M. L. Fernald has, as usual, been a constant source of aid and inspiration. The authorities at the National Herbarium of Canada have rendered great assistance, in these days of limited travel, by loaning their large collection of willows from the region in question ; and Dr. H. A. Gleason, Curator of the New York Botanical Garden, very kindly gave me access to the collections in his charge. ARTIFICIAL KEY TO THE SPECIES BASED ON WELL-DEVELOPED FLOWERING OR FRUITING SPECIMENS A. Bracts of the pistillate flowers not persistent in fruit, straw-colored; stamens 2,.0.00% more; aments coetaneous or serotinous, borne on leafy peduncles; usually rather tall shrubs of alluvial shores, or occasionally of muskeg swamps .........SER1Es I (p. 89). A. Bracts of the pistillate flowers (at least toward the apex) brownish, ,yellowish-brown, fuscous, or atrofuscous, more rarely straw-colored, persistent in fruit; stamens 1 or 2: aments appearing before or after the leaves, sessile or pedunculate. B. Pistillate flowers with two glands, ventral and dorsal to the pedicel; prostrate, repent shrubs; filaments (unknown in S. jejuna) more or less pilose ......SErtEs II (p. 89). B. Pistillate flowers with only one gland, ventral to the pedicel; repent or upright shrubs; staminate flowers with one or two glands; filaments pilose or glabrous. C. Low, prostrate shrubs with rooting branches, or very small suffruticose plants ; aments coetaneous or very often serotinous; the leaves mostly well-developed at anthesis. SeriEs III (p. 89). C. Erect or arborescent shrubs, with branches never repent and rooting (though many species are of very low and spreading form) ; aments usually appearing before or With ThE LEAVES 6s cc ccc ci edeasevecvevevcceccsceeeseeseeee ce SERIES EV (p. 90). 1943] RAUP, WILLOWS OF HUDSON BAY REGION AND LABRADOR 89 — _ _ SERIES | . Petioles provided, toward the upper end or at the base of the blade, with distinct, often irregularly lobulate glands; leaves distinctly and rather closely serrate-dentate; aments mostly coetaneous but sometimes serotinous, borne singly on foliolate peduncles; ovary usually on a distinct pedicel 2-4 times longer than the gland; stamens 3 or more in each flower. 2. Mature leaves distinctly whitened or glaucescent beneath, the blades elliptic-lanceolate and rather short-acuminate at the apex; aments often appearing late in the season, short-cylindric, scarcely twice longer than thick; capsule 7-12 mm. long, on a pedicel ADOUt:HhIaS LOMS cates see aaa een fe. seas a aelicles a ACS he ba Se een eR cote rot 1. S. serissima, . Mature leaves green on both sides, though often somewhat paler green beneath, broadly ovate, elliptic, or elliptic-lanceolate, more or less abruptly caudate-acuminate; aments coetaneous, frequently 2.5-3 times longer than thick; capsule 4.5-6.5 mm. long, with a relatively longer PeGiCel: pac sece oro is sisi toca wi ds meta ae tm ease, Foals 2.05 /uctad: po . Petioles or bases of the blades without glands; leaves linear or linear-lanceolate, sub- entire or rather distantly and often subspinulosely serrate ; aments usually serotinous, some- times appearing 2 or 3 on the same peduncle or lateral branchlet; capsule short-pediceled, the pedicel hardly more than twice as long as the gland; stamens 2.......... 3. S. interior. SeriEs I] . Aments mostly very short, and borne laterally on the branchlets; capsules 3-4 mm. long ; lateral veins of the small, rounded, reticulate-veined leaves, even those near the base, clearly rising from the midrib; rare, prostrate species from near the Straits of Belle Ai | eR AR la, 6d oo Sy o. o 6 on (0) Ae) Sndvionn’'a a,'40 9 e14 anwiinss, and ool aUOLtg ny ERED N eles 6. S. jejuna. . Aments borne pseudo-terminally ; capsules 4-7 mm. long; lateral veins of the leaves as in the last, or fewer, with two or three pairs rising from near the lower end of the mid- rib; leaves prominently reticulate-veined, rather coriaceous in texture. 2. Prostrate shrubs, with leaves glabrous or nearly so; lateral veins of the leaves few, the lower ones rising from near the base; petioles elongate (up to 3 cm.) ; capsules scarcely over 4.5 mm. long; floral bracts only short-pilose on the inner surface ...4. S. reticulata. 2. Shrubs with prostrate or erect branchlets; leaves usually conspicuously silky-pilose on both sides (or occasionally glabrate), the veins well-scattered along the midrib; petioles not much longer than the buds; capsules 5-7 mm. long; floral bracts more or less silkcy-pilose on. both sides... .. scjp ce eee reas es ess secs ss cemeaserganh O.. Vesitta, Series III . Ovaries and pedicels glabrous, even when young; leaves mostly crenate-dentate or glan- dular-crenate-denticulate (or usually entire in S. arctophila f. lejocarpa). 2. Leaves of the same color on both sides, green, usually very small, crenate-dentate, usu- ally rounded or often cordate at the base, sometimes retuse at the apex; floral bracts more or less uniformly colored, yellowish or violet, sparingly short-pilose; tiny arctic plants with semi-herbaceous stems and short, few-flowered aments .....8. S. herbacea. . Leaves distinctly paler beneath, often pruinose. 3. Ovaries with distinct pedicels, even when young, the pedicels of the capsules nearly always 2-3 times longer than the glands; adult leaves entire or at most inconspicu- ously glandular-denticulate or crenate, never marcescent. 10. S. arctophila f. lejocarpa. 3. Ovaries subsessile, or the pedicels, even in fruit, somewhat shorter or scarcely longer than the glands; leaves glandular-crenate-denticulate (at least toward the middle and apex), the mature ones marcescent, partly long-persistent; prostrate, matted shrubs with tiny leaves; common on the Labrador Peninsula ............... 7. S. Uva-ursi. bo . Ovaries (sometimes only in part or only the pedicels) more or less densely pilose ; capsules sometimes glabrescent; leaves entire or at most inconspicuously glandular-denticulate or crenate. 4. Bracts of two colors, fuscous at the apex or over a large part of the surface, more or less long-silky toward the apex, sometimes only ciliate at the apex. 5. Ventral glands rather short and broad, scarcely twice as long as thick, broadly trun- cate at the apex, 14 the length of the pedicels of the fruit; common species of arctic habitats, with prominent, upright, reddish aments .................. 10. S. arctophila. 90 SARGENTIA |4 5. Glands oblong, mostly 214-4 times longer than thick; pedicels of the fruit shorter than the glands or rarely a little longer. 6. Leaves ovate or ovate-elliptic, oval or obovate-oblong, often plicate at the apex, rounded to broadly cuneate at the base, glabrous or nearly so, green and shining above, resembling those of S. arctophila, though somewhat narrower; rare species of the west and northwest coasts of Hudson Bay .............. 11. S. hudsonensis. 6. Leaves variously shaped, from elliptic-obovate to broadly rounded-ovate, green above, but scarcely shining, silky-hairy at least when young; a common and highly variable species as represented in Eastern Arctic America ......9. S. arctica vars. 4. Bracts uniformly colored, straw-colored, yellowish, or light brown, rather short-pilose (the hairs shorter than the bracts), sometimes glabrate on the inner surface. 7. Petioles very short, rarely up to 2 mm. long; rare, prostrate species of northern Hudson Bay and the arctic continental coast ............ 00000 15. S. fullertonensis. 7. Petioles well-developed, mostly over 2 mm. long; the more prostrate forms of the NN WR a tg ao. 26:99 £6 oe 5. plc ony bee ee 13. S. cordifolia. Series 1V 1. Ovaries and pedicels glabrous, evey when young. 2. Pedicels distinct, even at anthesis 2-6 times as long as the glands; styles (and stigmas) short. 3. Bracts uniformly colored, yellowish or yellowish-brown, in life sometimes a little reddish, often becoming brown in drying; aments coetaneous, even the flowering ones borne on foliolate peduncles. 4, Leaves giving off a balsamic odor, even long after they are dried; leaflets and leaves glandular-crenate-serrate all around, very often cordate at the base, broadly ovate, elliptic, or elliptic-oblong; petioles 5 mm. long or more; glands short, somewhat thicker than long; capsules rostrate, 7-9 mm. long exclusive of the pedicels. 21. S. pyrifolia. 4, Leaves not giving off a balsamic odor; leaflets and leaves entire, not cordate, nar- rowly lanceolate to broadly oblanceolate, elliptic-lanceolate, or elliptic; petioles, even of the mature leaves, scarcely more than 6 mm. long; glands oblong-ellipsoid ; capsules rather obtusely conic, 6-7 mm. long exclusive of the pedicels; erect shrubs, 1-3 ft. high, of bogs and wet shores ............30. S. pedicellaris var. hypoglauca. 3. Bracts of two colors, fuscescent for the most part or at the apex; flowering aments subsessile on the branchlets. 5. Styles 1-1.25 mm. long; aments at anthesis 2.5-3.5: 0.9 cm., in fruit up to 8:2 em. (exclusive of the peduncles) ; capsules 9-10 mm. long exclusive of pedicels 2-2.5 mm. long; upper leaf surfaces bright green and more or less shining; shrub up to 16 ft. high, of gravelly shores and alluvium ............... 19. S. glaucophylloides. 5. Styles hardly over 0.8 mm. long; aments and capsules shorter ; upper leaf surfaces rather dull green. 6. Flowering twigs more or less yellow or flavescent; capsules 4-5 mm. long ex- clusive of pedicels 1.5-3 mm. long; tall shrubs of river flood plains ...17. S. lutea. 6. Flowering twigs fuscous, brownish or castaneous. 7. Leaves glaucous beneath; medium to tall shrubs of wet soil ..... 16. S. cordata. 7. Leaves of the same color on both sides (except in myrtillifolia var. brachypoda) or only a little paler green beneath; low, spréading shrubs of mossy bogs. 18. S. myrtillifolia. 2. Pedicels none or short, even in fruit’ not more than twice as long as the glands; styles (and stigmas) varying in length. 8. Stipules of the preceding year persistent; aments thick-cylindric, sessile, 1-3 pseudo- terminal toward the ends of short twigs; branchlets, at least when young, densely hirsute-villous or woolly. 9. Stipules linear- to half-cordate-lanceolate, acuminate, glandular-serrate-dentate ; twigs of the year hirsute-tomentose; leaves obovate-elliptic or elliptic-lanceolate or obovate-oblong; petioles 5-15 mm. long; mostly upright, medium shrubs of the northwestern, part Of OUT area con... csccescccccccvveveveudes 22. S. Richardsonii. 9. Stipules half-cordate to reniform, entire or glandular-denticulate; twigs of the year hirsute-lanate ; leaves broadly ovate or elliptic-orbicular, often cordate at the base: petioles commonly not over 2, or scarcely ever over 4 mm. long; mostly low, erect, wpresditig OF mintted shrubs: ic oo ose sis eda dandvacecdcnwisess 23. S. calcicola, 1943] RAUP, WILLOWS OF HUDSON BAY REGION AND LABRADOR 91 8. Stipules none or not persistent ; aments often coetaneous, sessile or pedunculate, for the most part distinctly lateral. 10. Aments, even in flower, distinctly pedunculate, provided at the base with small, rather normal leaflets; pubescence of the bracts scarcely or not at all curly. 11. Leaves distinctly glandular-serrate with fine teeth, often fimbriate, more or less Gensel y DUO ic Fike odes eck a gs acne nee Soe eee ee pies N eset Os adéenopnylia. 11.” Leaves entire, sparsely floccose-lanate .....0 562 Fowens wees «- - 24. S. Wiegandii. 10. Aments, even in fruit, sessile or subsessile, provided at the base only with very small, scale-like leaflets; pubescence of the bracts more or less distinctly curly. i2>-Medium: to tallsshriibsof low, eround! 22.7 Sete sees eee iia 16. S. cordata. 12. Low, spreading shrubs of mossy bogs ...... 18a. S. myrtillifolia var. brachypoda. 1. Ovaries (rarely only the pedicels or the base of the ovaries) pilose. 13. Pedicels distinct, even at anthesis 2-6 times as long as the glands. 14. Bracts uniformly colored, yellowish-straw-colored or yellowish-brown; aments coeta- neous. 15. Styles none or very short, not longer than the short-oblong stigmas; pubescence of the leaves villous-tomentose or the leaves glabrescent; medium or tall shrubs, com- monly in dry or damp woods and thickets ..........0.02000000+++-29, S. Bebbiana. 15. Styles distinct, equaling or somewhat longer than the stigmas. 16. Pubescence of the leaves silvery-silky; flowering twigs glabrous, fuscous; cap- sules, exclusive of the pedicels, 4.5-5.5 mm. long; low, erect or bushy shrubs of Arctic’ and al Pille: MADItAtS os oe tere sic a eo cw ee seiea ns won 40. S. argyrocarpa. 16. Pubescence of the leaflets and flowering twigs villous-tomentose; capsules about 7 mm. long, exclusive of the pedicels; tall shrubs, 10-14 ft. high, of Newfoundland atid thes Maingate lslanus wie dee wives eineioy. cass s aseeeomn tere eeeeee 28. S. cryptodonta. 14. Bracts of two colors, fuscescent for the most part or toward the apex; aments often precocious. 17. Stigmas rather long, 4-5 times longer than thick, mostly 1-1.5 (-2) mm. long, longer than the styles which are often obsolete; capsules 7-14 mm. long exclusive of pedicels 1-3.5 mm. long; large shrubs of low, wet ground .......33. S. discolor. 17. Stigmas short-oblong, scarcely over 3 times longer than thick, usually less than 1 mm. long; styles evident, about as long as the stigmas; capsules usually shorter than in the last; medium shrubs of dry woods and barrens .........32. S. humilis. 13. Pedicels none or short, even in fruit scarcely more than twice as long as the glands. 18. Flowering twigs more or less distinctly pruinose; aments precocious; pubescence of the leaves distinctly lustrous; tall shrubs of swamps and alluvium ....... 36. S. pellita. 18. Flowering twigs not pruinose, or if so the leaves glabrous; pubescence of the leaves if present dull or silvery-silky. 19. Bracts of the flowers uniformly colored, yellowish-straw-colored or yellowish- brown. 20. Petioles very short, mostly 1-2 mm. long, not exceeding well-developed buds; stipules about equaling or twice as long as the petioles. 21. Prostrate shrubs with rooting branches; rare species of northern Hudson Bay and the continental arctic coast; aments in fruit cylindric, 2:1 to 4: 1.3 cm. 15. S. fullertonensis. 21. Small upright shrubs with branches often rather short and somewhat twisted ; aments: short-oblong to spherical Gees... oes ee ees 14. S. brachycarpa. 20. Petioles well-developed, longer than the buds or stipules. 22. Leaves usually much longer than broad, those of the peduncles oblong, about 2.5-3.5 times longer than wide; aments usually loosely flowered toward the base. 23. Leaves subcoriaceous, glabrous or glabrate (sparsely pilose when young), closely resembling those of S. pedicellaris var. hypoglauca ..31. S. simulans. 23. Leaves not coriaceous, silky-pubescent when young and often rather tardily glabrate. 24. Styles very short, sometimes divided nearly to the base, much less than 1 mm. long; plants of the Hudson Bay region, mostly low or medium shrubs. : 12. S. glauca. 24. Styles distinct, 1 mm. long; rare species of N. Newfoundland; shrubs, up to 6.or 7 ft. high 2... scion asics os vee elea ee 37. S. amoena. 22. Leaves usually short-oblong or -ovate to rounded, those of the peduncles broader and shorter, scarcely up to 2.5 times longer than broad; low, spreading, or matted shrubs