ry REPORT OF THE CANADIAN ARCTIC EXPEDITION 1913-18 VOLUME V: BOTANY PART B: CONTRIBUTIONS TO THE MORPHOLOGY, SYNONYMY, AND GEOGRAPHICAL DISTRIBUTION OF ARCTIC PLANTS By THEO. HOLM OTTAWA F, A. ACLAND PRINTER TO THE KING'S MOST EXCELLENT MAJESTY 1922 Issued February 10, 1922 IONIAN Report of the ee Arctic Expedition, 1913-18. hg VOLUME I: GENERAL INTRODUCTION, NARRATIVE, ETC. "Sau Ae NORTHERN PARTY, 1913-18. By Vilhjalmur Stefansson.-........00seeeees _. (In preparation). Part By ‘SOUTHERN PARTY, 1913-16. By: Rudolph Martin Anderson.........++ee06+ Gs id iaadi a ie VOLUME II: MAMMALS AND BIRDS wus tests Part A: MAMMALS OF WESTERN ARCTIC AMERICA. By Rudolph Martin oem See J ons spas Bs ain’ aay Uae sc hgaanagBOS ciapb es aGos ghia espe esti duand basbsna dathehtoeeo anh X soeng RAM Brige e eee Te n preparation Part B: BIRDS OF WESTERN ARCTIC AMERICA, By R. M. Anderson and P. A. Taverner. : (oS Am scgcizeyaneseganaeduncteacroe eyo sien sso dacaifenaedetecd ate ans bal tbs We anaterbsshayzneviuarsvany ene i coatate poop inl aeepinae a | : Say : - ; VOLUME Ml: INSECTS a ae ; : yi : ryckooporon: By, C. Gordon Hewitt.. sas abaias iad lar cata S viderdee issacabere (Issued Fdsuiber 10, 1980). ” Part A: COLLEMBOLA. ‘By Justus W. Folsom..,............0eeeeeceeeeeeeeee .. Issued July 10, 1919). Part B: NEUROPTEROID INSECTS. By Nathan Banks mace wes R ceaintts tplase wiaieys (Ussued July'11, UE -, Part C: DIPTERA. , : os al ; “" Crane-flies. By Charles P, Mexender, } a. OF oi Mosquitoes. - By Harrison G. Dyar. : Diptera (excluding Tipulide' and “Galicia By. eal Melos or ee iene July aoe nie Part D: MALLOPHAGA AND ANOPLUR. : Mallophaga. By A. W. Baker. ~ : Anoplura. By G. F. Ferris and G. HB. F. Nuttall icy waoeeues enee cigs -ssued September 18 1919), Part B: COLEOPTERA. S Forest Insects, fable Ipide, Cerambycide, and Buprestides. By J. M. Swaine, ioe Carabide and Silphide. By H.C. Fall. : ; Coccinellide, Elatéeride, Chrysomelide ‘and Rhyachophorm, (excluding pide). CO. W. Le ng. Dytiscide. By J.D. Shermian, Jr......... 2.00. so asses: 08? ‘ aciare Dabtoee 12,,1919).~ Part F: HEMIPTERA. By Edward P. VanDuzee!. Big .- Ussued July 11, ‘1919).. Part G: eat ta ‘AND PLANT GALLS.‘ Z Saiwflies. - (Tenthredinoidea). By Alex. D . MacGillivray. : gre “e ‘Parasitic Hymenoptera. By Charles T.. Brues. ‘ oe re Wasps and ae By F. W. L. Sladen. Plant Galls. ‘By E. Porter Felt.............cc see eee sere e cece wes _(Tooud Ni ovember 8, 1919). Pat H: SPIDERS, MITES, AND MYRIAPODS. * ¥ ) Spiders. By J. H. Emerton. 3 — ‘ dey : _. Mites. By Nathan Banks: a hi . Myriapods. By Ralph V. Chamberlin...... Sshune wae aeedate nice wetbige ae (Issued ‘Tuly 14, 1918), Part I: LEPIDOPTERA, By Arthur pisen. Peek ely ea eee renee teen ene nen ery . Issued January 10, 1920). Peed, ORTHOPTE By: Bs M; Walk tics sates shes oars ems s cess Heroes (Issued September 4. 1920), Part K: IN aeT LIF ON THE WESTERN ARCTIC coast OF Asti. By Frits Johnsen. , ' se “VOLUME IV: BOTANY _ Part A: FRESHWATER ALGAR, AND FRESHWATER DIATOMS. By Charles Ww. Tow, : sasha: sion: Polcae ee ale ahcs dna ita tee oh biglad elnan sea ced aetna eatin bees revsidva ayavgsontouaacssddessuscosscavacl elt n preparation). Part B: MARINE ALGAE. By ‘F, Ss. Goilins. sie Wisid Hass Ste “1.0.1... ([n preparation), Part C: FUNGI. By John Dearness........... yateeeeeees é ',... Un preparation). Part D: LICHENS. By G. K. Merrill... ccc cece eee eee eect een e eee e nee ‘(In preparation). - Part E; MOSSES, , Ee R.S. , Williams Uiglgns wea cere dite done Bpprha Skis celeav Shs HR har vaerde’ (Issued February8, ial). : 8 eUME Vv: BOTANY a i gt Part A: VASCULAR. PLANTS. By James M. Macoun and Theo. Holm.. a Rent October 14, 1921}. - Part B: CONTRIBUTIONS TO THE.MORPHOLOGY, SYNONYMY, ‘AND GEOGRAPHICAL - DISTRIBUTION OF ARCTIC PLANTS. By Theo. Holm................... (In press), Part C: GENERAL NOTES ON ARCTIC VEGETATION. By Frits Johansen. (In preparation). : ; 7 » VOLUME VI: FISHES, TUNICATES, ETC. | “lart re FISHES. By F.Johansens.......cccccsgeesesesecseecenecs | ;.(In.preparation) » Part B: ? ABOIDIANS, ETC. By A. G. Huntsman. ecacnsnisas ie ..-. (Zn preparation). aes ; VOLUME Vi: CRUSTACEA PartA: DECAPOD. CRUSTACEANS. By MaryJ.Rathbun............. oxy sClnendd Asapuat 48, 1918). Part B: SCHIZOPOD CRUSTACENAS. By Waldo L. Schmitt. i i (Ussued September 22, ei. : Part C:CUMAGEA. By W.T.Calman........5......ccceeeeeeeet speck eeeeeees. Issued October 16, 1920). Part D:ISOPODA. By P. L. Boone................000ee ere ahd tise a .3.. (Issued November 10, 1920). Part E: AMPHIPODA. By. eres. = Shoemaker:.............005 (Issued September ry 1920). ‘Part. F: PYCNOGONIDA. LeonJ. Cole.........000¢eeee eee Bs . Issued January 8, 1921). Part G: EUPHYLLOPODA. By F. daar 4 ‘ si thsc have wtenoresatork (In preparation). Part H: CLADOCHRA. a ae eieedi eee eT ee sdtansieasditertatoesa - (Issued-June 28; 1920). Part I: OSTRACODA, By R. W. Sharpe...... 0. een ceeceece cee eeeeeeeeeteeeeeeetenens (In preparation). . Part J: FRESHWATER are ag Cc, Dwight ion sated ataveasdians (Issued April 21, 1920). Part K: MARINE COPEPODA. | By A. Willey.........: .... Pie 88 se Daas ws. Ussued June 26, 1920). Part L: PARASITIC COPEPODA.: By Charles B. Wiison... OR Se a (issued August 6, 1920). Ms 4 Part M: CIRRIPEDIA, By ‘Hi A. ae eye eee e eee ee podvbtaxseastbtet +++ (dn preparation). REPORT OF THE CANADIAN ARCTIC EXPEDITION 1913-18 VOLUME V: BOTANY PART B: CONTRIBUTIONS TO THE MORPHOLOGY, SYNONYMY, AND GEOGRAPHICAL DISTRIBUTION OF ARCTIC PLANTS By THEO. HOLM OTTAWA F, A, ACLAND PRINTER TO THE KING’S MOST EXCELLENT MAJESTY 1922 24657—1 Issued February 10, 1922. Cornell University Library The original of this book is in the Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924074095435 Contributions to the Morphology, Synonymy, and Geo- graphical Distribution of Arctic Plants. By Tuo. Hom, Clinton, Maryland, U.S.A. With six photographs and eighteen figures in the text drawn by the author. The present contribution is divided into four chapters: I, containing notes on morphological characteristics and synonomy; II, giving the geographical distribution; ITI, concluding remarks, and IV, the bibliography relative to the distribution. The fact that the collection made by the expedition contains many interest- ing species, together with the fact that they have been collected and prepared with great skill and care, has enabled me to examine their various organs, principally the vegetative ones, and Iam therefore able to offer some descriptions of ramification, reproduction, hibernation, etc., of which several points are but little known from arctic plants. Having had the opportunity myself to see the arctic flora, (Nova Zembla and Greenland), and moreover to compare this with the alpine flora (Rocky mountains, Colorado), I naturally feel induced to treat both as far as the scope of the subject will permit. Therefore, in the chapter dealing with the distribu- tion, I have inserted several columns for alpine plants, although the arctic distribution has been given the most extensively, because, in respect to geograph- ical distribution, the arctic and alpine floras are so intimately connected with each other that a discussion of either one alone would give very little information about their history. Considering together the interesting chapters on geographical distribution in Darwin’s ‘Origin of Species’ and Nathorst’s ‘‘Polarforskningens Bidrag till Forntidens Vaxtgeografi,” we have a most valuable foundation for further studies in this line, and quite especially with reference to the arctic flora. CHAPTER I. MORPHOLOGICAL CHARACTERISTICS AND SYN- ONYMY. GRAMINEAE. The species collected belong to the following tribes: Phalarideae, Agros- tideae, Aveneae, Festuceae, and Hordeae; of these the Festuceae are the best represented. They all are perennial and, concerning the habit, the stoloniferous type is somewhat more frequent than the caespitose one; widely creeping stolons above ground are characteristic of Glyceria vilfoidea; subterranean ones of Poa arctica, Arctagrostis, Dupontia, Festuca rubra var., Elymus, Arctophila, and Alopecurus; in the last three genera the stolons attain quite a considerable length and ramify freely. The culms are always simple and usually short, especially so in Glyceria tenella and G. vilfoidea, while in Elymus, Arctophila, Arctagrostis, and Dupontia, the height of the culm may reach forty cm. or even a little more. The inflorescence is most often an open, lax-flowered panicle, notably so in Arctophila, Dupontia, and Poa arctica; a contracted, spike-like inflorescence occurs in Trisetum, Alopecurus, and Calamagrostis; a spicate one in Elymus, Agropyrum, and Hordeum. The flowering glume ' is more or less hairy 1The old, well known term ‘‘flowering glume’ has recently become substituted by ‘‘lemma”’ proposed by Professor C. V. Piper (Science, N.S. vol. XXIII, 1906), and is introduced in various manuals. Not speaking of the fact that ‘‘lemma’’ has been in use for several years before as an anatomical term, proposed by Strasburger, it seems unwise to change the old term ‘‘flowering glume,” since this glume (gluma florens) as well as the empty glumes (glumae vacuae) are bracts of the same order, borne on the same rhachis, whereas the ‘‘palea” is borne upon a rhacheola, developed from the axil of the flowering glume. More- over, in speaking of ‘‘lemma’”’ in plurals, it is absolutely incorrect, from a linguistic point of view, to write ‘“Jemmas’” instead of ‘‘lemmata.”’ 3 24657—14 4B Canadian Arctic Expedition, 1918-18 in Elymus, Poa abbreviata, Festuca rubra var. arenaria, etc.; itis awned in Alope- curus, Trisetum, Festuca, Bromus, and in Hordeum the awn reaches a considerable length, up to about four em. In habit these arctic Gramineae do not exhibit any characteristics which might indicate the extreme conditions under which they live. The same observation I have made with respect to the alpine species, notably in the Rocky mountains, Colorado, several of which do not recur in the lowlands. Even the peculiar genus Pleuropogon with its rather large, drooping spikelets forming a simple, loose raceme, and of which one species,’ P. Sabinit R. Br., is almost circumpolar, is also represented in California by P. refractum Benth. and P. californicum Benth. As compared with the species known from the arctic American archipelago ! there are only a few types which were not found by the Canadian arctic expedition, for instance: Hierochloa alpina, Agrostis, Deschampsia, Catabrosa, and Pleuropogon, some of which are widely distributed in the arctic regions. But since Mr. Johansen succeeded in finding several of the small and much more inconspicuous species of Carez, there is no reason to believe that the “missing’’ species of grasses were overlooked. Arctagrostis latifolia (R. Br.) Griseb. I have without hesitation referred all the specimens collected by the Cana- dian expedition to this species although those from Camden bay, Herschel island, and Bernard harbour do resemble A. arundinacea (Trin.) Beal, as far as ' the panicle being more open and the spikelets less compact. According to Grisebach in Ledebour’s Flora Rossica (p. 434) A. arundinacea is considered merely a variety of A. latifolia: ‘‘panicula laxiori, palea inferiori (the flowering glume) obsolete 5-nervi.” Recently, by Scribner and Merrill,? this variety is enumerated as a species with no other distinctive characters than “the branches of the panicle being fasciculate, spreading or ascending,”’ besides some measure- ments of the culm, leaf, and spikelets. However, judging from the large material which I have examined from the northern regions of both Worlds, there seems to be no doubt of the genus being monotypic, although very variable. Specimens from Siberia, viz: Jenisei, Dudino, thus represent the typical plant as well as the variety, and they were found growing associated with each other; in several specimens of gigantic size from Jakutsk, Siberia, the panicle is very open and the spikelets comparatively small, but the structure of both the empty and the flowering glumes shows no deviation from that of the typical plant; similar specimens from Yukon and St. Paul island, Bering sea, differ in no respect from the type except by their greater size, the more open panicle, and the smaller spikelets. And in this respect Arctagrostis agrees with species of several other genera, for instance with Deschampsia caespitosa, as represented in the alpine region, the arctic, and the lowlands much farther south; furthermore, in Festuca ovina and in many species of Poa the panicle may be more or less contracted and the spikelets varying considerably as to size and number of flowers. Poa abbreviata R. Br. and P. glauca M. Vahl. Small forms of P. glauca may sometimes be mistaken for P. abbreviata. Lange? describes both, calling attention to the characters as follows: P. abbreviata R. Br. “Humilis, densissime caespitosa; foliis rigidiusculis, subsetaceo-involutis, acuminato-pungentibus, margine scabriusculis, ligula brevi, obtusiuscula; pani- cula brevi, simplicissima, subsecunda, ramis in verticillis c. 2 plerumque spiculam solitariam, 3-5-floram gerentibus; glumis acutis, inaequalibus; gluma florens (palea inf.) obovata, trinervia, puberula et inferne longius villosa, obtusissima v. truncata, eroso-denticulata.” 1Simmons,H.G. A Survey of the Phytogeograrliy of the Arctic American Archipelago. (Lunds Univers. Arsskr. Afd. 2. Bd. 9. Lund, 1913.) .2 Grasses of Alaska, 1910. * Conspectus Florae Groenlandicae (Medd. om Gré¢nland, Kj@benhavn, 1880). Arctic Plants: M. orphology and Synonymy 5 B _ The accompanying figure A: 3 I have drawn from a specimen collected in Spitzbergen by Professor A. G. Nathorst. 1 2 5 4 6 M, 2322 (js | 24 25 26 lil, Y Figure A. 1. Poa glauca M. Vahl; spikelet. (Wollaston land). 2. Same; flowering glume, side-view. 3. P. abbreviata R. Br.; spikelet. (Spitzbergen). 4. Glyceria vilfoidea (Ands.) Th. Fr.; empty glume. (Greenland). 5. Same; flowering glume, side-view. 6. G. maritima (Huds.) We.; empty glumes. (Norway). 7. Same; flowering glume, dorsal view. 8. G. Vahliana (Liebm.) Th. Fr.; empty glumes. (Greenland) 9. Same; flowering glume, palet and flower. 10. G. tenella Lge.;spikelet. (Martin point, Alaska). 11. Same;empty glumes. 12. Same; flowering glume, side-view. 13. Same; apex of flowering glume. 14. G. angustata (R. Br.) Fr.; empty glumes. (Spitzbergen). 15. Same; flowering glume and palet, side-view. 16. G. paupercula Holm; empty glumes. (Hudson bay). 17. Same; flowering glume, side-view. © 18. G. vaginata Lge.; empty glumes. (Cape Bathurst). 19. Same; flowering glume and palet, front-view. 20. Same; flowering glume and palet, side-view. 21. G. Kjellmanni Lge. ; empty glumes. (Nova Zembla). 22. Same; lower empty glume, dorsal view. 23. Same; upper empty glume, dorsal view. 24. Same; flowering glume, ventral view. 25. Same; a stamen. ‘26. Same; the pistil with the lodiculae. (All the figures are enlarged.) P. glauca M. Vahl: “Caespitosa, multicaulis, rigidiuscula, magis minusve glaucescens, spithamaea v. parum ultra; culmis laevibus (P. caesia) v. superne scabriusculis (P. aspera), articulo superiore elongato; foliis subdistichis, planis v. complicatis, apice curvato cucullato-contractis, ligula brevi, obtusa, folii supremi ovali-oblonga; panicula erecta, angusta, ramis brevibus, ante et post anthesin adpressis, scabris; spiculis 3-5-floris, lanceolato-ovatis ; glumis sub- aequilongis, acutis, violaceis, albomarginatis, gluma florens (palea inf.) 5- nervia, ad nervos sericea et basi lanata. Formae maxime memorabiles | sunt: 6 elatior Ands. 1-2’ altus, foliis planis; paniculae ramis longioribus magisque quam in ‘ a effusis, c. 3-5 in verticillis. : ; ; ifn ; ; “‘y pallida Lge. Dense caespitosa, pallide straminea, caespitibus basi vagin- arum anni praéteriti reliquis cinctis; foliis angustissimis, complicatis, ligula lacera; spiculis majusculis, lanceolatis, 5-6- floris; gluma florenti et palea lividis, albomarginatis, cum macula aurea sub apice, ad nervos sericeis, ceterum glabris. 6B Canadian Arctic Expedition, 1913-18 “§ atroviolacea Lge. Spithamaea, rigida, fol. planis, caulino superiore patulo basin inflorescentiae sub-attingente, ligula elongata, panicula virginea dense coarctata, spiculis subbifloris, glumis longe acuminatis, atroviolaceis, gluma florente et palea basi viridibus albomarginatis, apice purpureo-marginatis.’”’ 5 ae spikelets (Fig. A: 1-2) are from a typical specimen collectedon Wollaston and. ; In P. abbreviata the broad empty glumes as well as the flowering glume, the latter being puberulent on the sides and along the keel, constitute a good distinction, when compared with P. glauca, in which the glumes are much narrower, and the flowering glume hairy only on the keel. P. arctica R. Br. There is no doubt about this being the same as P. flerwosa Wahlenb., but it is debatable whether it is to be considered identical with P. cenisia All. Lange (l.c. p. 178) states for instance: ‘Nomen P. cenisie All., quod a pluribus autt. hue relatum est, non ad hanc speciem spectare videtur, si cum Gren. et Godr. ad P. distichophyllam Gaud., a nostra bene distinctam, P. cenisia ut synonymon ducitur. “‘P_ arctica R. Br. vero ex consensu plur. autt. omnino eadem cum nostra; utrum horum nominum praeferri debet, incertum videtur, cum uno eodemque anno (1824) publicata sunt.” As pointed out by Lange, (l.c. p. 178) P. arctica is readily distinguished by the spreading, flexuose branches of the short panicle; by the ovate, 2-4- flowered spikelets of which the glumes are purplish and of which the flowering glume is obtuse or roundish at apex, with a broad, hyaline margin, villous at the base, and pilose along the veins. The rhizome is stoloniferous. Arctophila Rupr. The old genus Colpodium of Trinius! was founded upon two species: mon- andrum and Stevent, which by Trinius himself were considered as “‘species facie dissimiles,’’ and they are indeed so unlike that Robert. Brown 2 a few years later segregated the former as Phippsia algida R. Br. and retained the latter only as a Colpodium. To the latter genus was furthermore referred C. latifolium R. Br. although Robert Brown was not certain about the real affinity of this species to those of Trinius and especially not to C. Steveni and compressum. At present Robert Brown’s species latifoliwm is generally placed under Grisebach’s genus Arctagrostis as A. latifolia (R. Br.) Griseb. While Colpodium of Trinius was adopted by Grisebach *? with the omission of C. monandrum (Phippsia), the genus was nevertheless augmented with certain species placed as a section “‘Arctophila” in contrast to C. Steveni and its natural allies representing the section ‘“Hucolpodium.’’ By including the species of Arcto- phila Rupr., the genus Colpodium became actually an aggregate of incongruities, as it had been before with Phippsta and Arctagrostis included. By Bentham the genus was finally restricted to the section Eucolpodium, while Arctophila became transferred to Graphephorum Desv., next to Glycerta R. Br.; another disposition was made by Hackel who followed Grisebach by placing both Eucolpodium and Arctophila as sections of the original genus Colpodium, characterizing them as having the “spikelets one- to two-flowered, etc.’”’ This same classification is also followed by Beal ‘ in his Monograph of North American Grasses, with the .same erroneous characterization; erroneous, because it was originally intended for Colpodiwm alone in the sense of Trinius. 1 Trinius: Fundamenta Agrostographia. 1820. p. 119. 2 Brown, Robert: Chloris Melvilliana. 1823. (Miscell. bot. works, I, p. 221). 3 Ledebour: Flora Rossica. IV. 1853, p. 384. 4 Beal, W. J.: Grasses of North America, II. New York, 1896, p. 556. Arctic Plants: Morphology and Synonymy 7 8B Three species are enumerated by Beal as representatives of the genus Colpodiwm in North America: C. fuluwm (Trin.) Griseb., C. pendulinum (Laestad.) Griseb., and C. mucronatum (Hack.) Beal. Considering the fact that Colpodium in the sense of Trinius was originally intended for both C. Steveni and Phippsia algida, it seems difficult to find any good ground for admitting species of so little affinity as those of Arctophila and still crediting the genus to Trinius. And the species of Arctophila have themselves been transferred’from one genus to another. Thus we find them as members of Poa, Glyceria, Graphephorum, and finally of Colpodium. ; Considered by themselves, the species of Ruprecht’s Arctophila' constitute an excellent little genus, and we might cite Ruprecht’s own words when he proposed the genus in his “Flores Samojedorum cisuralensium”: _ “Arctophila a Catabrosa (airoide) praesertim differt glumarum conforma- tione et longitudine, hac nota etiam et insuper valvulis ecostatis a Glyceria R. Br. recedit. Atropis Trin. (P. distans) Catabrosae quoad glumas proxima, spiculas habet (saltem in’ statu virgineo) lineares, fere teretes; in Arctophila nostra semper ex ovato-oblongae vel lanceolatae. Conjunctioni Arctophilae cum Poa obstant: valvulae dorso concavae vel saltem minus compressae; flosculi lana numquam cincti, nec ad nervos dorsales sericei, sed ad callum more Aven- acearum pilis rigidis obsiti; valvula inferior apice vix integerrima, sed margo plerumque irregulariter denticulatus et erosus, saltem crenulatus et apex saepe obtusus vel truncatus; habittis etiam nobilior colore fulvo paniculae saepe intermixto; spiculae majores plerumque et flosculi demum patuli, remotiusculi.” The species that are best known are: Arctophila fulva (Trin.) Rupr., A. pendulina (Laestad.) Ands., and A. effusa Lge., especially the first of these since the Greenland plant, A. effusa, was for many years considered identical with A. pendulina by Fries, Grisebach, and several other authors. The species found by Mr. Johansen is A. effusa Lge., described in Conspectus Fl. Groenl. as follows: ‘“‘Satis superque differt planta groenlandica, in Fl. Dan. tab. 2343 nomine Poae pendulinae divulgata, a Glyceria pendulina Laest. vera, Lapponiae tornensis et rossicae incola, statura humiliore foliis brevioribus, culmo erecto (nec apice nutante), ramis deflexis, in verticillo 1-2, raro 3, flosculis in spicula 2-3 (nec 3-6) minus laxe dispositis, glumis obtusiusculis, spicula parum brevioribus, etc. Botanici plures (I. Vahl, Fries, Ledebour, etc.) has 2 species, inter se valde dissimiles, infauste junxerunt, quare plantam Groenlandicam l.c. nomine novo designare coactus sum.” West Greenland, between 64° 10’ and 65° 20’, in damp situations, very rare; Sukkertoppen, Godthaab (Vahl). A depauperate form of this species is known from Spitzbergen, formerly considered as distinct and named Colpodiwm Malmgreni Ands.? While Grisebach (l.c.) only recognized A. fulva and A. pendulina, seven other species had been described by Ruprecht (1.c.) but merely referred to as synonyms in Flora Rossica; they were collected on the island Kolgujew and on the Russian coast near Kambalnitza and Bjelaja. The genus seems to be rare in North America, but specimens belonging to it have been collected in various parts of Alaska and adjacent islands and also in Canada. These specimens have been generally identified as A. fulva or A. pendulina; of these the former is known only from Muckelung river and the Hudson bay region, and, so far as I know, A. pendulina has never been found on this continent. In the copious material sent to me for identification from the Canadian Government, I found four species, which appeared distinct, and of which three have been described and figured in my paper: On the genus Arctophila Rupr.’ 1 Ruprecht, F. J.: Symbolae ad historiam et geographiam plantarum Rossicarum. 1846, p. 62-65. 2 Anderson, N.I. Bidrag till den nordiska Floran I. Ett hittills obeskrifvet gras fran Spetsbergen. (Ofvers. Kongl. Vet. Akad. Férhdlgr. No. 5, p. 121, Stockholm, 1866.) 3 The Ottawa Naturalist, June, 1902, p. 83. 8 B Canadian Arctic Expedition, 1913-18 The species A. gracilis nob. from north of Lake Superior, collected by Dr. R. Bell; A. brizoides nob. from St. Paul island, Bering sea, by James M. Macoun, and A. chrysantha nob. from sixteen miles west of Nome city, Alaska, by J.B. Flett. Furthermore A. trichopoda nob. from Mansfield island and Nottingham island, Hudson bay, collected by Dr. R. Bell, and of which the diagnosis ! reads as follows: “Stoloniferous: culms weak, about twenty cm. in height, enclosed by the leaf-sheaths: leaf-blades broad, flat and glabrous, longer than the internodes: inflorescence a lax-flowered panicle with the glabrous, capillary branches in twos and threes: spikelet three-flowered, one or two on each branch: empty glumes unequal, quite broad, acuminate, membranaceous, three-nerved: flower- ing glume broad, pointed and often awned, at least in the basal flower, hairy at the base: palet bidentate, shorter than the flowering glume. A near ally of A. mucronata Hack.” Finally may be mentioned that John Murdoch collected an Arctophila near Point Barrow which Hackel has described as A. mucronata, fide Beal: Grasses of North America.? The diagnosis reads as follows: “A smooth, stout grass, 15-25 em. high. Leaves 6-8 in number crowded; ligule broad, lacerate, 2 mm. long; blades flat, abruptly pointed, 5-12 cm. long, 6 m.m. wide. Panicle shining, yellowish, open, partially included, narrow or pyramidal, 5-7 mm. long, rays in twos or fours, reflexed, the longest 3-7 cm. long, bearing 2-3 spikelets near the apex. Spikeléts 2-flowered, joint of rachilla O,6 m.m. long, smooth or very sparingly hairy; empty glumes subequal, 3-3, 3 m.m. long, soft, thin, first ovate, 1-nerved, second broader, 3-nerved; floral glume broadly oval, 3,5 m.m. long, 3 mm. wide, 5-nerved, margin scarious, apex irregularly toothed or torn, the central nerve extending to the tip or into a short mucro; palea broad; 1,5 mm. long.” It is thus characteristic of A. trichopoda and A. mucronata that the midrib of the flowering glume is extended into an arista in the former, and in a mucro in the latter, while in all the other species the midrib does not extend beyond the apex of the glume, a structure which certainly would be anomalous in the genus Colpodium Trin. (as understood by Bentham), because there the midrib never reaches the top of the glume. These species of Arctophila may naturally be classified in two sections: I. Macrostachyae. scar de when fully developed, five-to seven-flowered, the base acute during anthesis. A. fulva, A. remotiflora, and A. pendulina. II. Brachystachyae. Spikelets two- to four-flowered, the base obtuse during anthesis. A. brizoides, A. chrysantha, A. deflexa, A. gracilis, A. latiflora, A. mucronata, A. poecilantha, A. scleroclada, A. similis, and A. trichopoda. Dupontia R. Br. The genus was established by Robert Brown ° and characterized as follows: “Gluma subaequivalvis, scariosa, concava, mutica, locustam 2-3-floram sub- aequans. Perianthia mutica, scariosa, (basi barbata), altero pedicellato; valvulis integris, inferiore concava. Lodiculae 2. Ovarium imberbe. Stigmata sub- sessilia. Caryopsis— Gramen glabrum erectum. Folia linearia, plana, vaginis semifissis, basi integra. Panicula simplex, coarctata, fusco et purpur- ascenti varia, pedicellis cum locustis continuis, perianthiis separatim solubilibus.”” The genus was named in honor of Monsieur Dupont of Paris, author of a valuable essay on the sheath of the leaves of grasses, and of observations on the genus Atriplex. 1 Fedde’s Repert. l.c. 2Vol. 2, New York, 1896, p. 556. *Chloris Melvilliana l.c., p. 228. Arctic Plants: M orphology and Synonymy 9B Only one species, D. Fischeri, was known to Robert Brown, but since then Ruprecht (1.c.) has described a second one, D. psilosantha, from Kolgujew island, and a third one, D. micrantha nob., has been found in Labrador: Halton, and at cape Henrietta Maria, Hudson bay. fe Of these D. psilosantha differs from D.Fischeri by the empty glumes being acutissimae, caudatae,” and by the flowering glume being “(flosculi) glaberrimi, acuminati,” while in D.Fischeri the empty glumes are more or less obtuse and the flowering glume obtuse and pilose at the base; the third species, D. micrantha, differs from both by the small size of the mostly one-flowered spikelets, etc. according to the diagnosis? “Stoloniferous: culm slender, glabrous, about twenty-five cm. in height: leaves very narrow, glabrous with prominent ligule: inflorescence a contracted panicle with capillary branches: spikelets one or two on each branch, mostly one-flowered: empty glumes unequal, very narrow and sharply pointed, mem- branaceous: flowering glume acute, slightly hairy at the base, not exceeding the see empty glume: palet bidentate, a little shorter than the flowering glume. Glyceria R. Br. The segregation of Atropis originated with Ruprecht? and was accepted by Grisebach in Flora Rossica: “genus a Glyceria imprimis stylo ad basin usque simpliciter plumosa distinctum est.’”” Meanwhile Parlatore* established the genus Puccinellia containing the same species as Atropis and formerly by Elias Fries‘ referred to his section Heleochloa of Glyceria. According to Fries (Le. 1846) Glyceria consists of two sections: ‘‘ Hydrochloa Hartm.,” with the flowering glume 7-nerved, and “ Heleochioa Fr.” with the flowering glume 5-nerved. As a section Aéropis has been recognized by various writers, and much more so than Puccinellia. Buchenau *, however, treatsGlyceria and Atropis as distinct genera, and this author is one of the very few who furnishes a sufficiently com- plete diagnosis. In Gray’s Manual (1908), Puccinellia is accepted as a genus, though poorly defined as distinct from Glyceria; more recently we find the same disposition by Fernald and Weatherby (Rhodora 1916), enumerating eleven species as indigenous to eastern North America south of Hudson straits. However, in giving Puccinellia preference to Atropis Fernald and Weatherby have certainly misunderstood Ruprecht, because this author has clearly demon- strated that he considered Atropis and some other genera as being distinct from Poa. And when Ruprecht states that: “E conditione glumarum generum series fortasse sequens: Dupontia, Arctophila, Poa, Atropis cet,’ this author did not mean that these: “from the condition of the glumes perhaps represent a series of genera as follows: Dupontia cet.,” but that the genera mentioned might be arranged according to the structure of the glumes, as enumerated above; series means in this connection sequence referring to the arrangement. More- over, it seems unjust to ignore Atropis because the diagnosis of Puccinellia by Parlatore is more complete; if such procedure be considered ‘‘the best for serving the cause of sound nomenclature,” as claimed by these authors, very many genera of Linnaeus would suffer the same fate. By comparing the species of Glyceria, the classification proposed by Fries (l.c.) appears the most natural, retaining the genus intact, as has been proposed also by Duval-Jouve in his classic treatise: Doutes et priéres au sujet de quelques espéces de Glyceria du groupe des Halophiles.® 1 Fedde’s Repert, l.c. 2 Florae Samojed. Cisural. 1846. 3Fl. Ital. I. 1848. 4 Summa veget. 1846. 5 Flora d. nordwestdeutsch. Tiefebene, 189!. 6 Bull. Soc. Bot. France; Paris, 1863. 10 B Canadian Arctic Expedition, 1913-18 The species enumerated in the following pages represent a certain type of the genus, and the only one known from the arctic regions which by some authors has been considered worthy of generic rank, distinct from Glyceria sensu strictiori. As a section of Glyceria this type was described already by Elias Fries' under Heleochloae, while most of the other species were referred to the section Glyceriae genuinae Kunth. According to Fries the principal charac- ters of the latter were “Stylus distinctus, pilis stigmatis denticulatis. Valvula exterior septemnervis, ‘nervis elevatis, etc.’’ In the Heleochloae the character- ization reads: “Stigmata subsessilia, simpliciter plumosa. Valvula exterior 5-nervis, nervis obsoletis. Festucae spec. Kunth.” : Very brief and incomplete is the diagnosis by Hackel,? namely: “Glycerta, ae distinct, lodicules grown together; and Atropis, styles none, lodicules istinct.”’ Glyceria vilfoidea (Ands.) Th. Fries (Fig. A: 4-5). This well-known and, on the arctic shores, so very frequent species has recently been transferred from one species to another or reduced to a mere variety. By Scribner and Merrill it is considered identical with Trinius’ Poa phryganodes (1.c.) and has suffered a similar fate in the hands of Fernald and Weatherby (l.c.). It would, however, seem very strange if the earlier authors familiar with the species described by Trinius had not discovered that Anderson’s Catabrosavilfoidea was identical with it. Trautvetter,? for instance, does not cite the species of Trinius as a synonym of Glyceria vilfoidea (Ands.); neither do Kjellman and Lundstrém in their extensive writings on arctic botany, nor Lange in his Conspectus (l.c.). As long as we feel sure, and indeed unquestionably so, about the identity of Glyceria vilfoidea but not of the other plant, it would naturally be more safe to retain this name, even if it be of a younger date than Poa phryganodes. Dr. Trimen laid down a very good canon about names in botany, viz.: to take the most certain name, even if it be of a younger date; for nothing works so badly as to take an uncertain name; it is disputed, and we may have several names running at once. In recent years works dealing with enumerations of plants from various countries are actually filled with new combinations because the authors believe that they have detected a much older name than the one now in use; very many of these old names, however, are so uncertain that the earlier writers discarded them, and the result is, of course, the introduction of an older name, which is supposed to be correct, until still another be proposed, thus involving an immense amount of new combinations. Examples of this kind abound in recent works; we need only mention the much disputed Poa triflora Gilib., Carex fusca All., C. Halleri Gunn., etc., names that were never accepted by the earlier writers on systematic botany. And as long as the old masters discarded such names, they surely must have had some reason for doing so; in any case the earlier writers, I mean Trautvetter, Blytt, Hartman, Fries, Lange, Wahlenberg, for instance, were certainly more familiar with the writings of Gilibert, Allioni, and Gunner, for instance, than authors of to-day. Recently G. vilfoidea has also been referred to G. maritima as a variety, respectively ‘“Atropis” or “‘Puccinellia,” thus involving “new combinations.” Meanwhile Krok‘ claims that G. vilfoidea was actually discovered in the year 1838 by Laestadius in Finmarken, and that he called it G. distans var. reptans Laest., but without publishing the diagnosis. The fact that the plant became described by Hartman ® several years before Anderson described his Catabrosa vilfoidea, induced Krok to suggest the name G. vilfoidea to be changed to G. reptans (Laest.). However, since the name G. vilfoidea is the one used by the leading authors on arctic botany, and since we know the species with absolute certainty, the name “‘vilfoidea”’ proposed by Anderson ought to be retained. 1 Novit. Florae Suecicae. Mant. II, 1839. 2in Engler’s: Die nat. Pflanzenf. 3 Incrementa Florae Phanerogamae Rossicae. Fasc. IV. Petropolis, 1884, p. 850. 4 Tvanne i Finmarken Aterfunna fanerogamer (Bot. Notis. Lund, 1899, p. 137). 5 Excursions Flora Fd.I. 1846. Arctic Plants: Morphology and Synonymy ll B _ _ To consider the species as a mere variety of G. maritima, as has been done in recent years, is open to question, and Johanson ! has written a very instructive paper, demonstrating that G. vilfoidea and G. maritima are not conspecific. Lange gives a good diagnosis of G. vilfoidea (Conspectus|1.c. p. 170) as follows: Gracilis, glaberrima, rhizomate obliquo, foliorum fasciculos-abbreviatos et flagella valde elongata gracilia emittente; foliis anguste linearibus subcompli- catis, acutis; ligula brevi, truncata; panicula brevi, contracta, ramis sub anthesi adpressis, defloratis erecto-patentibus, inferioribus subgeminis, super. solitariis; glumis obtusis inaequalibus flosculis 2-3, palea inf. ovali, obtusa, 3-5-nervia.” “Obs. Species haec, non nisi e Groenlandia et ins. Spidsberg. hucusque cognita, modo crescendi aliisque characteribus a G. maritima recedit, etsi hujus formis quibusdam G. maritima var. arenaria E. Fries Mant. 2. p. 9) affinis est.” Ficure B. 1. Stolon of Glyceria vilfoidea (Ands.) Th. Fr.;P =the prophyllon (fore-leaf); L=lateral shoot; S=primary shoot; natural size. (St. Paul island, Bering strait.) 2. Stolon of Catabrosa aquatica (L.) Beauv.; St =flower-bearing stem; other letters as above; natural size. (Sweden). 3. Two viviparous spikelets of Aira alpina L.; magnification 13 (Greenland). For a comparison I have drawn the spikelets of both (Fig. A: 4-5 and 6-7). Moreover, as pointed out by Johanson (l.c.), the structure of the stolons differs in a marked degree from that of the other species of Glyceria, and seems indeed to be rare, or at least but little known from Gramineae in general. The accompanying text figure (B) shows part of a stolon of G. viifoidea (fig. 1); in this the main shoot is aerial (S), and bears several green leaves, separated from each other by stretched internodes; at each node a small lateral shoot is developed (L), of which the first leaf represents an adorsed prophyllon (P) or fore-leaf, as is the usual structure of lateral shoots in the monocotyledons. But, as may be seen from the drawing, the lateral shoot is not situated in the axil of the corresponding stem-leaf, but has been moved up to the node above, thus occupy- ing a very strange position, remote from the leaf-axil. It is a structure that recurs in the genus Catabrosa, for instance, in C. aquatica (L.) Beauv., where the structure is very plainly visible (fig. 2); in this plant the stolons appear to root more freely, the secondary roots proceeding from the nodi, as shown in the drawing. The fact that G. vilfoidea seldom produces flowers, and that fruiting speci- mens are almost unknown, indicates that vegetative reproduction supersedes the floral, and in this respect G. vilfoidea is readily distinguished from the other species of the genus. While vegetative reproduction is known to be prevalent 10m Glyceria maritima (Huds.) Wahlenb. var. arenaria och Glyceria vilfoidea (Ands.) Th. Fr. (Sv. Bot. Tidskr. 4. Stockholm, 1910). 12 8B Canadian Arctic Expedition, 1918-18 in several other plants, notably arctic and alpine, the organs of reproduction in these are flowers transformed into small shoots or bulblets, instead of being stolons developed from the basal leaves. In Aira alpina, for instance (fig. 3), the spikelets are transformed into leafy shoots, since the flowering glume, the palet, and the sexual organs have developed in the shape of green leaves with sheaths and blades, while the empty glumes show the typical structure of glumes. A similar transformation of flowers into leafy shoots is known also from the arctic-alpine Polygonum viviparum, from some of the Saxifragae, S. cernua and 8. stellaris, where most of the flowers develop in the shape of bulblets, but the almost total absence of flowers and the vegetative reproduction being plese by means of true stolons seems to be especially characteristic of Glyceria vilfoidea. G. tenella Lge. and G. vaginata Lge. were also collected on the expedition and since the former has been so very poorly described in Ostenfeld’s Flora Arctica, and the latter having been merged into G. distans (L.) Wahlenb. by this - same author, I deem it necessary to insert the original diagnoses of these, and of some other arctic Glyceriae, which may prove useful to future students of arctic plants. Some of the species have been illustrated with respect to the structure of the spikelets in the text-figures. Glyceria tenella Lge. (Fig. A: 10-13) is described in Kjellman and Lund- strém’s Fanerogamer fran Novaja Semlja: “Gl. caespitoso-pulvinata, multi- caulis, caulibus gracilibus circa 3” longis, adscendentibus (v. prostratis?); foliis angustissime linearibus, complicatis, obtuse et oblique mucronatis; ligula brevissima, truncata; panicula brevi, laxa, racemiformi, contracta, florendi tempore e vagina folii caulini superioris vix exserta, ramis laevibus 1-2 in verti- cillis, inaequilongis; spiculis glabris, circa 3-floris, floribus invicem remotis; glumis obtusiusculis, superiore subduplo longiore; palea inferiore violacea vel viridi, anguste albomarginata, obtusa vel subtruncata, leviter erosa, 3-5 nervi; antheris pallide fuscis vel sordide luteis; ovario elliptico-ovali, stylis invicem subdistantibus terminato. Hab. In sinu Rogatschew insularum Novaja- Semlja et ad promontorium Grebeni insulae Wajgatsch legerunt F. R. Kjellman et A. N. Lundstroem mense Julii 1875. “Habitu nec non pluribus characteribus a reliquis Glycerits mihi notis dis- tincta, maxime ad Gl. vaginatam Lge. et Gl. distantis var. pulvinatam accedens, cum nullis harum tamen associanda.”’ G. tenella Lge. forma pumila Lge. (in Holm: Novaia Zemlia’s Vegetation). “Differt a typo in insulis Novaia Zemlia a cl. Kjellman et Lundstroem lecto: statura pumila, culmis vix ultra 1’ longis, panicula(racemo brevi) florendi tempore extra folium sup. exserta; palea inf. magis obsolete nervata. Forsan G. tenella ipsa (hac forma inclusa) ad G. vaginatam Lge. e Groenlandia (Fl. Dan. T. 2583) ut forma reducta trahenda est, sed haec panicula magis effusa et ramosa, spiculis 5-6 floris gluma inf. latiore differt. Hab. in rupibus insulae Novaia Zemlia ad Petuchowskoj Schar. Glyceria vaginata Lge. (Consp. Fl. Groenl.) (Fig. A: 18). “Dense caespitosa; culmo ad basin 1-2 nodis geniculato, articulo superiore elongato, fere ad paniculam usque vagina ampla, folii lamina multo longiore incluso; foliis mollibus, anguste linearibus, laevissimis, glaucis, planis vel sicca- tione involutis, ligula brevi, obtusa; panicula laxa, ramis subgeminis, capillaribus, ante et post anthesin erectis; spiculis laxe 5-6-floris, glumis ovatis, obtusis, inferiore duplo breviore, palea inf. concava, obtusissima, obsolete 5-nervia, basi pilosiuscula, super. paulo longiore, truncata, apice eroso-ciliata.’”’ Known. so far only from the coast of West Greenland. Glyceria? Kjellmanni Lge. (Fig. A: 21-26.) (in Kjellman & Lundstrém: Fanerog. fr. Novaja Semlja). Arctic Plants: Morphology and Synonymy 138 _, ‘Gl. perennis, caespitosa, multiceps; fasciculis singulis caespitis polyphyllis; foliis inferioribus ad vaginas reductis, foliis caulinis 1-2, omnibus planis, laxi- usculis, e basi lata sensim angustioribus et oblique acuminatis, ligula protracta, acutiuscula, leviterque denticulata; culmis erectis vel geniculato adscendentibus, 2-5 longis; panicula e folio caulino supremo longe exserta, brevi, angusta et -coarctata, ramis 1-2, brevibus, erectis cum rhachi laevibus et glaberrimis; spiculis circa 2: floris; glumis ovatis, inferiore pellucida, 1: nervi, superiore 4 longiore, 3: nervi et circa nervos violaceo-tincta, margine pellucida lacera; palea inferiore ovato-lanceolata, nervis 3, latis, violaceis 2 longitudinis parte percursa, basim versus dorso et margine dense pilosa, apice irregulariter denti- culato-lacera, palea superiore margine revoluta ad nervos 2 pilis albis, rigidis ciliata apice summo denticulata; antheris pallide fuscis vel sordide luteis; ovario elliptico stylis 2, invicem approximatis terminato. “Hab. In insulis Novaja Zemlja ad fretum Matotschkin legerunt Kjellman et Lundstroem, 7-13 Julii 1875. . “Habitum Glyceriarum minus bene exprimit et obiter inspecta potius Pots quibusdam similis, sed characteribus, imprimis glumis dorso teretibus a Poae genere recedit, nec reliquae characteres obstant, quin ad Glyceriae genus referri possit, nisi quod spiculae fere semper bi-florae et paleae dense villosae, superior immo margine rigide ciliata (fere ut in Bromo). Habitu Dupontiis haud dissimilis est, sed hoc genus glumis spicula aequilongis praeter pluraabunderecedit. Itaque hoc gramen singulare ad interim Glyceriae generi subsumsi, nisi forte proprium genus constitueret.” Glyceria Vahliana (Liebm.) Th. Fries. (Fig. A: 8-9). “Poa Vahliana (Liebm.): gramen caespitosum spithamaeum, radice fibrosa, culmo 3-foliato, vagina superiore folio 14 pollicari longiore, ligula producta oblique truncata; panicula contractiuscula radiis binis solitariisque, spiculis subtrifloris pedicellis longioribus vel aequilongis; glumis subaequalibus-elongato- lanceolatis obtusiusculis glabris obsolete-nervosis, basi apophysi parva instructis; paleis glumas, superantibus valvulis inaequilongis, exteriore lineari-lanceolata obtusiuscula obsolete 5 nervia, nervis pilosulis, interiore breviore apice bidentata nervis 2 marginalibus purpureis ciliatis percursa, lodiculis oblique bidentatis, germine elliptico obtuso, stigmatibus 2 plumosis.” Flora Danica, 41, tab. 2401; Kjoebenhavn, 1845. The species was found by Vahl on the west coast of Greenland, by Niakornak near Umanak, at an elevation of about 700 M. Lange (Conspectus Fl. Groenl.) adds to the diagnosis: “foliis flaccidis, planis, acutissimis; glumis parum inaequalibus, purpurascentibus, margine apicem versus albomarginatis.” Glyceria Langeana Berlin (Karlvaxt., insaml. under svenska expeditionen till Grénland 1883). “Humilis, dense caespitosa; foliorum fasciculis intravaginalibus; foliis rigidiusculis arcuatis; culmis rigidiusculis vix folia superantibus; paniculis et spiculis glabris; glumis inaequalibus acutiusculis; paleis obtusis exterioribus apice laceratis.”’ North Greenland: Kangaitsak. Glyceria angustata (R. Br. sub Poa) Fr.1(Fig. A: 14-15). By Robert Brown (Chloris Melvilliana,p.223) the speciesis described as follows: “Poa angustata, panicula simplici coarctata lineari-lanceolata, locustis 4-5 floris, gluma inferiore dimidio minore, perianthiis apice erosis: valvula inferiore basi elanata lateribus glabriusculis, foliis angusto-linearibus. “Desc. Gramen 4-6-unciale, glabrum, erectum; radice fibrosa. Culmi foliati, basi quandoque divisi, laeves. Folia angusto-linearia, plana, acuta, glabra, laevia; vaginae subcylindraceae, laeves, suprema folio proprio longior, 1 Novit. Fl. Suec. 3.Mantissa Lund, 1842. 4B, Canadian Arctic Expedition, 1913-18 omnes ipsa basi integra; ligula subquadrata tam lata quam longa, apice dentata dente medio paulo longiore. Panicula erecta, angustata, circumscriptione lineari-lanceolata, ramis paucifloris, pedicellis denticulatis, strictis, viridibus, apice paulo dilatato, cum locusta haud omnino continuo. Locustae hyalinae, glaberrimae, uninerviae, cum pedicellis persistentes, valvula inferiore fere dimidio minore; superiore duplo latiore et fere duplo longiore, obtusiore, peri- anthio dimidio circiter breviore, nervis lateralibus obsoletis. Perianthia separ- atim decidentia, rachi locustae glabra; valvula inferior oblonga, concava, acutius- cula, apice scarioso eroso-denticulato, quinquenervis, lateribus infra medium pube rara in nervis extimis crebriore instructis, ipsa basi absque lana implexa; superior paulo brevior, dinervis, nervis viridibus, denticulatis, lateribus com-~- plicatis. Lodiculae 2, hyalinae, imberbes, semibifidae. Stamina 3.” According to Lange the species is a near ally of G. Vahliana. Glyceria paupercula Holm (in Fedde, Repert. III, 1907) (Fig. A: 16-17). “Caespitose, perennial: culms low, curved or decumbent, mostly leafy to the panicle, the base of which is often enclosed in the leaf-sheath; leaves narrow, flat, glabrous, the blade about as long as the sheath; panicle pyramidal and open during anthesis, more or less contracted afterwards; spikelets three to four-flowered, seldom five-flowered; empty glumes unequal, the lower the shorter, both much shorter than the spikelets; flowering glume pointed, carinate, glabrous, five-nerved; palea bidentate, glabrous. “It is a member of the section Atropis Rupr. ‘Mansfield island, Hudson bay; collected by Dr. R. Bell.” Since I described this species a revision of the genus Puccinellia in eastern North America has been presented by Fernald and Weatherby, in which several new stations have been recorded from the coasts of Labrador, Keewatin, New- foundland and Quebec, besides that two varieties have been described by these authors: alaskana (Scribn. et Merrill), new comb., and longiglumis n. var., the latter from Prince Edward island. According to the revision, cited above, the var. alaskana has been passing very generally in eastern America as Puccinellia angustata, based upon Poa angustata R. Br., as for instance, by Simmons (Vascular plants, Ellesmereland, 1906), whose specimens thus belong to the variety of paupercula. Glyceria arctica Hook. (Fl. Bor. Am. II, p. 248.) (Fig. C,). Figure C.—Glyceria arctica Hook. (Greenland). 1. Spikelet. 2. Empty glumes. 3. Flowering glume, side-view. 4. Same, dorsal view. 5. Palea. (All figures enlarged.) Arctic Plants: Morphology and Synonymy 15 5 Lange (Consp. Fl. Groenl. 1.c.) describes this as follows: _ , “Caespitosa, 1-13’ longa, culmis e basi geniculato adscendentibus v. erectis, rigidis (v. rarius laxiusculis); foliis planis, exsiccando involutis; panicula ampla (ad 6” longa), rachi inflorescentiae scabro, ramis firmis, erecto-patulis (raro divaricatis), ante et post anthesin adpressis; spiculis majusculis 5-8-floris, glumis acutiusculis, infer. duplo breviore; palea inf. 3-5-nervia, obtusa, apice eroso- denticulata. “Var: ‘8 laxa nob. G. arctica Hook. ex descriptione) Dur. pl. Kan. No. 97 (ex loco natali) foliis latioribus, planis, flaccidis, pagina superiore scabris; paniculae ramis infer. 2-3 in verticillis, post anthesin divaricato-reflexis; spiculis varie- gatis, 6-8-floris. “y capillaris nob., culmis flaccidis; foliis elongatis, angustis, supra laevibus; panicula nutante, ramis longiusculis, capillaceis, flexuosis 2-3 in verticillis, de- floratis erecto-patulis, inferne longo spatio nudis, spiculis laxe 5-6-floris, viridibus. § dasyantha nob., rigidiuscula, praecedentibus humilior, palea inf. e basi ad medium dorso lateribusque pubescente, basi, (Poaruwm instar) longe lanata. “Species polymorpha, praecedenti G. marit.) affinis et forsan non nisi ejusdem varietas, tamen modo crescendi, statura elatiore et pluribus characteribus videtur distinguenda. Nomen G. arcticae Hook. adhibui, cum descriptio (l.c. p. 248) satis bene nostrae plantae respondeat; observari tamen debet figuram (Bor. Am. tab. 229) evidenter ad specimen macrum et nondum plene evolutum factam esse. Suppositionem (in textu ad Fl. Dan. tab. 2582), plantam groenlandicam eandem esse cum G. maritima v. palustri Fr. adhuc bene fundatam esse censeo.”’ Greenland. ‘ Elymus mollis Trin. This species is frequently confounded with the arctic E. arenarius L.var. villosus R. Mey. By Grisebach (FI. Ross. l.c.) H. mollis is described as follows: “Culmo apice velutino, foliis demum convolutis, spica elongata, spiculis geminis v. ternis 5-plurifloris rhacheos internodium superantibus, glumis late lanceolatis acuminatis 5-7-nervibus margine membranaceis floribusque dense villosis, his exsertis, nervis tenuibus dorso prominulis.”’ The variety villosus of E. arenarius is described by Lange (l.c.): “Humilior quam forma in Europa vulgaris, sed spica robustior, glumis dense villosis; ceterum vix a type differt.”’ CYPERACEAE. Carex [Ruppius] L. The genus is poorly represented on the arctic coast explored by the expedition and C. incurva is the only Vignea collected. With regard to the habit the stoloniferous species are more frequent than the caespitose, and they all are “‘phyllopodae.” This character “phyllopoda”’ implies a biologic peculiarity which is possessed especially by northern types; it consists in the shoot being dicyclic, i.e. vegetative in the first, floral in the second season, with the rami- fication of the shoot sympodial. Moreover, in this type the very young inflores- cense is protected by being surrounded and covered by a number of leaf-sheaths. The inflorescence thus terminates the shoot and when the fruits have matured, the main shoot dies off completely while a lateral bud developed in the axil of one of the basal leaves continues the growth of the rhizome and repeats the succession of a vegetative and floral stage. Otherwise with the “aphyllopodae.” In these, principally southern species, the shoot is most often pliocyclic with reference to the vegetative stage but monocyclic as far as concerns the floral. For in the aphyllopodae the ramification of the shoot is most frequently mono- podial; however, there are some species which, although being aphyllopodae, show a sympodial ramification and in which the very condensed rhizome,mostly more or less caespitose, develops a number of purely floral and purely vegetative 16 B Canadian Arctic Expedition, 1913-18 shoots but with no apparent regular succession. Characteristic of the aphyllo- podae is, however, that the young inflorescence is not covered by leaf-sheaths from the year previous but by scale-like leaves or, let us say, “bud-scales developed at the same time as the bud appears. These types: phyllopodae and aphyllopodae, were first brought to attention by Elias Fries 1 and they are indeed very characteristic and helpful to the determination of many critical species. In my papers dealing with Cyperaceae? I have often demonstrated the distinction of these types as an important means of classifying Carices. With regard to the species, the list shows that the Vigneae are only repre- sented by C. incurva, while eleven Carices genwinae were collected. As has been the experience of other arctic expeditions, the Microrhynchae, Aeorastachyae and Physocarpae are the most frequent, and C. stans Drej. seems to be the most common; it represents C. aquatilis Wahlenb. of the more southern, but still boreal, regions. Since C. stans has sometimes been mistaken for tall specimens of C. rigida, I have figured one of the specimens from Herschel island (Part A, Plate I, figs. 1-4), beside a much larger one, which I collected on the west coast of Greenland, near Jacobshavn (Part A, Plate I, figs. 5-8). The original diagnosis * reads as follows: “Spica mascula 1, feminis 3-6, elongatis cylindricis densifloris obtusis subsessilibus, bracteis infimis foliaceis latis planis culmum superantibus auri- culatis, squamis oblongis obtusis, perigyniis (immaturis) obovatis brevissime rostellatis ore integris enerviis, stigmatibus 2-3. “Groenland; Vahl. “Radix stolonifera crassa vaginisobducta. Culmus erectus, strictus, crassus, obtusangulus v. acutangulus, laevissimus v. (raro) superne scabriusculus. Folia plana, lata, laevissima v. apicem versus margine scabrata. Spicae 4-8, terminalis mascula, reliquae femineae; mascula cylindrica subclavata squamis oblongis obtusis, femineae densiflorae obtusae, superior interdum apice mascula, infima pedunculata reliquae sessiles. Bractea infima longissima culmum superans, marginibus apicem versus scabris, auriculis brevissimis subamplectentibus. Squamae perigynio multo minores, obtusae, fuscae v. nigrae, v. perigynia superantes acutiusculae. Perigynia obovata enervia, estipitata brevissime rostellata ore integro, viridiflavae v. apice fuscae. Stylus brevis subexsertus, stigm. 2 v. 3 elongata, flaccida villosula. ‘““Hlaec species, cujus perigynia modo immatura vidi, ex distylis C. aquatilé et C. saxatili maxime affinis est. Ab illa differt foliis omnino planis latis, bracteis brevioribus, spica mascula solitaria, colore spicarum fusco v. atrato, statura humiliore. Ab hac: bracteis latis planis brevissime auriculatis culmum multo superantibus, spicis subclavatis densifloris cet. A ceteris distylis longius distat: C. acuta differt spicis pedunculatis pendulis acutis; C. caespitosa spicis acutis, bracteis angustioribus, culmo scaberrimo, perigyniis stipitatis nervatis. C. stricta Good. spicis magis laxifloris acutis vaginis foliorum fibrilloso-reticulatis, cet.” This diagnosis was based upon specimens collected by Vahl in Greenland. Kiukenthal ‘ refers C. stans to C. aquatilis as a variety including C. variabilis Bail., a species from the subalpine bogs in the Rocky mountains (Canada- Colorado); however, this classification is not natural, and is undoubtedly the result of working with too scant material. Concerning C. subspathacea and C. reducta, these are near allies of C. salina, but their geographical distribution is much farther north; they have been described and figured in a recently published paper, dealing with Carices aeoras- tachyae: Salinae.’ 1 Synopsis Caricum distigmaticarum, spicis sexu distinctis, in Scandinavia lectarum. Bot. Notiser. Lund. 1843, p. 97. 2 Am. Journ. of Science, Vol. 48. New Haven, 1919, p. 19. 2 Drejer, S. Revisio critica Caricum borealium. Naturhist. Tidsskr. Vol. 3. Copenhagen, 1841. 4 Cyperaceae-Caricoideae in Engler: Das Pflanzenreich. Leipzig, 1909. 5 Holm, Theo. Studies in the Cyperaceae. Am. Journ. of Sc., Vol. 49. New Haven, 1920. Arctic Plants: Morphology and Synonymy 17 B LILIACEAE. Bulbous ‘plants are extremely rare in the arctic region. Zygadenus glaucus Nutt. of the Melanthaceae, Lloydia serotina and Allium sibiricum of the Liliaceae are, so far as I know, the only bulbous species known from this region; they were collected by Kjellman at Port Clarence, while Lloydia is the only one collected by the expedition. The genus Liloydia has an interesting history and we owe to Irmisch! an excellent account of the history and of the morphological structure of the species. It was first described by Caspar Bauhin,? who named it Pseudonarcissus gramineo folio, sive Leuconarcissus aestivus; at that time it was known only from the mountains of Switzerland and Austria;.several years afterwards it was collected also in England, on the highest peaks of Snowdon‘in Wales by Ed. Lhwyd (also written Lloyd), and by Rajus designated with the uncertain appellation as: Bulbosa alpina juncifolia, pericarpio unico ereéto’ in,summo cauliculo dodran- tali. Rajus, as a matter of fact, had only fruiting specimens. By Dillenius it was named Bulbocodium alpinum; by Linnaeus Bulbocodium serotinum and later on Anthericum serotinym, until Salisbury referred it to a new genus which he named Lloydia. a Ae v “% . SALICACEAE. Salix Richardsonii Hook. - a The accompanying photograph (1) shows a tree of this species from Mackenzie river, south coast of Victoria island., Photograph (2) shows the same species, growing along a small tributary to Coppermine river (below Bloody fall). © * About the occurrence of this species in the river-bed of Sadlerochit river. Alaska, Mr. Johansen quotes from ‘his field-notes as follows: “Two species of willow-shrubs and trees- seem to occur in this river-bed and on the sheltered and wet places (tributary creeks) of the adjoining mountain slopes, about 25 miles from the coast: (1).A very common and dominant species (Salix Richardsonit Hk.) with reddish bark (branches) and pubescent twigs and leaves, the latter being rather broad and: ovate-shaped; in favourable places this willow attains more than man-height and’ a trunk thickness of up to about 6 inches in diameter and forms a dense tree or shrub growth.‘ (2) A not nearly sO common species (S. glauca L.) with more narrow leaves, and less pubescent twigs and darker bark. It does not attain (in these arctic river-beds) the size of (1), is hardly more than a shrub and has a far less twisted-forked appearance, but a more straight growth than (1). Also it seems to prefer more dry places than (1) and is first met with some distance from the coast inland. “Willows about man-height were seen on the low, gravelly clay banks or islands in the bed of the river at its mouth; but outside of these (and similar) protected places it was only in the larger creek-valleys that they attained a | similar or bigger size. Such a creek valley, bordered by 200-300 feet high clay banks on the east side of the river I went into. The clay banks had the usual tundra vegetation (including dwarfed willows) on top, and steep bare sides; but in the creek bottom willows had an extensive and luxurious growth (getting searcer and dwarfed at the head of the creek), attaining more than double man-height and trunks thicker than a man’s arm on specially protected places, (See Neg. 2). — ‘ x ¢ “There seemed to be two species, the one (S. Richardsonti Hk.) reaching a far greater size and being much more dominant than the other (S. glauca?). Similar conditions for and occurrences of willows were reported by different members of the expedition from other rivers (Tree and Hood rivers) flowing into Coronation gulf.” —— 1 Beitriige zur vergleichenden Morphologie der Pflanzen. Bot. Zeitung, Leipzig, 1863. p. 161. 2Prodromus. Basel, 1671. p. 27. 2 24657—2 18 B Canadian Arctic Expedition, 1913-18 Saliz Richardsonti Hook. A tree in creek-bed of Rae’s ‘‘ Mackenzie River,’’ south coast of Victoria Island ° 30’ _ March 29, 1916. (Photo by F. Johansen.) ictoria Island about 111° 30’ West. S. ‘anglorum Cham. The species was collected at several stations and, with regard to its occur- rence, Mr. Johansen writes: “Cape Bathurst. Grows among grasses, etc.,on the clay-banks with which the higher tundra behind falls off to the sandspit, and on the tundra itself, Found both on the top and slopes of these clay-banks; in the former place the vegetation is more stunted and scarcer than on the slopes and in the protected, small gullies here. Growth somewhat prostrate and spreading, but twigs and catkins protruding well from surrounding vegetation.” Arctic Plants: Morphology and Synonymy 19 B Saliz Richardsonii Hook. From tributary to Coppermine river (below Bloody Fall). February 10,1915. (Photo by F. Johansen. ) “Pihumalerksiak island, Dolphin and Union Strait, N.W.T. [off Cockburn point—R.M.A.]. The vegetation of this island is rather scarce and stunted Owing to its exposed position and poor soil (dolomite-rock or gravel); it is best developed in depressions, or in shelter of protruding cliff-parties or large boulders. The conditions and the vegetation are thus similar to those found on the island (except Chantry island) in the outer harbour at Bernard harbour. “The willows in question grew in patches here and there; generally several plants together; growth prostrate, and each plant not nearly as spreading as on the mainland. “Bernard harbour (mainland), Dolphin and Union strait. Very common everywhere, especially on poorer soil (sand, gravel), where it is one of the domin- ating plants. On such gravelly and sandy soil with little other vegetation and where more exposed the plant spreads out over the ground (not half buried in the sand as is the case with S. ovalifolia) in all directions, the whole plant remind- ing one of a huge basket-starfish. The centre of the plant is elevated most (thick trunk-bases, many dead leaves); and most of the leaves and catkins are found at the terminal branch-ends, which seem to seek out small depressions in the ground to shelter the twigs. The diameter of a very large plant may reach a couple of yards; its height over the ground only a few inches. As is the case with other prostrate willows the female catkins are less likely to be blown off the plants during the winter, where these latter (not the catkins) are snow- covered. In 1916 the season was several weeks earlier than in 1915, and the catkin-buds were noticed to emerge from their bracts in the end of May, 1916; though it took a whole month before the first flowers appeared. In 1915 the first staminate flowers were noticed June 23 and the first pistillate ones (and new leaves) five days later. The flowering begins earlier at the coast and on island than farther inland. The flowers in the catkins have the following colours: (1) staminate scales (bractlets) black with purple base and white hairs; small inner scale (bractlet) purple; filament pale-pink; anthers dark rose-purple, turning yellow when the pollen came out, later black. (2) pistillate: As the staminate (capsules) dark purple, below with white woolly hairs.” 24657—23 20 B Canadian Arctic Expedition, 1913-18 S. ovalifolia Trautv. var. camdensis Schn. About this Mr. Johansen writes as follows: “Collinson point (Camden bay). This willow grew on more bare, gravelly tundra near the beach (transition-region to the latter), in patches of several plants. Its growth was very prostrate and depressed (among stones and vege- tation) with the stems and branches lying very close to the ground and spreading widely, so that only the catkins showed up from a little distance. Especially the subterranean parts (roots and stem parts) were less extensive and spreading than with those found at Konganevik, Alaska (see below); probably because they, did not grow on sand dunes as is the case at the former place. “ Konganevik (Camden bay). The collecting place was where the seashore (beach) through low sand dunes goes over into the more typical tundra behind. On these sand dunes the vegetation is very characteristic and consists almost exclusively of Elymus, Carex, Salix, Chamaenerium, etc.; each species spreading (both above and under the ground) over large patches (areas) and dominating more or less to the exclusion of the other species. This Salix seemed to be very prostrate, but the larger part of each plant is buried in the sand, so that only the leaf- and catkin-carrying branch-parts (outer 3) protruded. It was mostly large plants widely spreading (both roots and stems); the branches often having form of long “runners” intersecting the sand in all directions. The sand-covered parts of the branches were without leaves or catkins and pale (white-yellow). When growing in less sandy soil the growth is naturally more condensed (see above under Collinson point). The plants were in full bloom in theendof June here. “Martin point. The collecting place was a sandy gravel spit of slight elevation with the sand dunes less pronounced than at Konganevik. Vegetation rather scattered and in patches, except around the several ponds and the big lagoon between the sand spit and the mainland behind. On sandy places the vegetation was much like that at Konganevik, with Honckenyia taking the place of Chamaenerium. As the character of the spit was somewhat intermediate between the beach regions at Collinson point and at Konganevik, so did also the growth of the Saltz in question resemble those of the species from both of the above places. At the time of collecting the plants had dropped staminate catkins and had unripe pistillate.” S. reticulata L. As may be seen from the text-figure (D) the foliage varies quite considerably as to size and shape; in nearly all the staminate plants collected the leaves were smaller than in the pistillate. The large, almost orbicular leaf (Fig. 1) is from a large pistillate plant of which all the leaves showed the same outline and approximately the same size; the two leaves (Figs. 2 and 3) are from a staminate plant and both forms of leaves occurred together on this plant; the specimens were collected at Bernard harbour. Ficure D. Leaves of Salix reticulata L. J. Pistillate plant;’ Bernard harbour. 2. Staminate plant; same locality. 3. Staminate plant; same locality. 4, Pistillate plant; Herschel island. 5 and 6. Pistillate plant; Bernard har- /( ee (All leaves show the natural {| size. ; Arctic Plants: Morphology and Synonymy 218 POLYGONACEAE. Polygonum viviparum L. Some tall specimens from Camden bay measure in height 16 cm. The subterranean stem is a short, but thick, horizontal rhizome with the internodes extremely short; towards the apex the rhizome is more or less bent so as to form a knee; no stolons are developed, and thin, secondary roots proceed from the rhizome. Characteristic of the species is the development of bulblets in the inflorescence; they have been described by Areschoug.! These bulblets consist of a tuberous axis of two internodes, and the leaves are membranaceous, sheath- ing; at the apex of the bulblet is a terminal bud enclosed within membranaceous leaves, mostly three. When the bulblet falls off a green leaf appears and second- ary roots develop from the tuberous internode. These bulblets thus form new individuals and constitute an important means for distributing the plant over a larger area since, as mentioned above, there are no stolons. i P. Bistorta L. Hjalmar Nilsson 2 has described the structure of the rhizome which accord- ing to this author represents a monopodium with a terminal, vegetative bud, and with the flower-bearing stem developed in the axil of one of the green leaves. The rhizome resembles that of the preceding species but it is stoloniferous and, on the other hand, no bulblets are developed in the inflorescence. The speci- mens collected measured a height of about 18 cm., including the inflorescence with the flowers well developed. Rumex arcticus Trautv. Several specimens were collected with flowers and mature fruit; the height of a fruiting specimen from Bathurst inlet measured 40 cm., the stem being 12 cm., while the large inflorescence was 28 cm. The subterranean stem represents a very condensed pseudo-rhizome borne upon a deep, thick tap-root; beside that, long and quite thick secondary roots develop from the subterranean internodes. Green leaves are quite numerous, forming a rosette. Oxyria digyna (L.) Hill. In this plant the aerial shoot represents a rosette of leaves with the flower- bearing stems apparently lateral. With respect to the subterranean organs there is a deep but relatively slender primary root which persists for several years; from the crown of the root several erect or ascending stems are developed; they are densely covered with remnants of leaves and the internodes are very short. From these subterranean stem-portions stolons develop (Fig. E); they are covered with tubular leaves but the internodes are often quite distinct, and secondary roots proceed from the nodi. The specimen figured represents a case of the aerial shoot being developed from a stolon, beside that a young stolon of second order has developed from the ascending, subterranean stem. Most of the specimens collected were of this type, developed from stolons, and only a very few were seedlings with the primary root present. CARYOPHYLLACEAE. Lychnis [Tourn.] L. sectio Wahlbergella Fries. As a genus Wahilbergella was established by Elias Fries* and distinguished by the ovary being unicellular, the valves of the capsule bifid, and the seeds winged; the genus was placed between Viscaria and Melandrium. 1 Bidrag tila Groddknopparnas Morfologi och Biologi. Lund, 1857, p. 23. Tab. VII, fig. 16. 2 Dikotyla jordstammar. Acta Univ. Lund. xix; Lund, 1882-83. 3 Botaniska Notiser. Lund 1843, p. 143. 22 B Canadian Arctic Expedition, 1918-18 This same classification we find in Fries’ Summa vegetabilium Scand- Stockholm (1846), with two species: W. apetala (L.) and W. affinis (J. Vahl), a diagnosis of the genus is given in this work (pp. 150-155): ‘““Capsula unilocu- laris, dentibus 10 per paria connatis dehigcens, carpophoro stipitato. Semina angulosa, limbo (in nostris saltim) tumido corrugato marginata. Reliqua Lychnidis (Melandryi) diversae capsulae dentibus simplicibus et aequaliter distantibus. Quoad capsulam differt ut Malachium a Stellaria; quoad semina ub Lepigonum ab Alsine. Omnes species arcticae, micranthae, caulibus simpli- cibus. That Fries also considered Melandrium triflorum (R. Br.) Vahl to belong to the genus may be seen from p. 155 (l.c.) where he writes: ‘“‘Proxima est W. triflorae Vahl seminum margine vix limbato diversae.”” While the majority of Scandinavian authors, Blytt, Hartman, Kjellman, Lindman, etc. have accepted Wahlbergella as a genus, Lange! placed the species under Melandriwm and recently Warming ? follows the disposition of Lange. Still another classification is proposed by Robinson ? placing Wahlbergella as a member of Lychnis Tourn. sectio Eulychnis Fenzl (extended), and this author is certainly correct when stating: ‘‘(Examination of a number of specimens seems to show that in American species the characters of partial septation of the capsule, division of the valves, inflation of the calyx, are very variable,and do not lead either individually or in combination to more definite or satisfactory results.” As long as the genus Wadlbergella was known only to consist of the two species mentioned above, W. apetala and affinis, it was quite natural, but when increased with Lychnis triflora B. Br. and furthermore with L. Taylorae Robins., L. montana Wats., L. Kingii Wats., and L. nesophila nob., the generic distinction vanishes. On the other hand, these species may naturally represent a section of their own: the arctic alpine Wahlbergella; it is interesting to notice that a member of the section, L. apetala L. var. gracilis has been recorded from the alpine region of western Tibet, alt. 15-17,000 feet. Th. M. Fries has described a variety arctica from Spitzbergen ‘ the diagnosis of which reads as follows: “‘pedunculis dense cinerascente, subviolaceo-glanduloso- villosis; calyce subgloboso-inflato, magis patente glanduloso, vulgo apice sang- uineo-violaceo, venis subnigricantibus longitudinaliter striato; petalis sub anthesi exsertis, violaceis (1. pallidis).’”” This variety was found on the coast of Wollaston land, as stated above. By the petals being very distinctly exserted this variety somewhat resembles L. nesophila nob.* from Mansfield island, Hudson bay, but differs with respect to the calyx being more inflated and also by the shape of the petals. Having studied various species of the sectionWahl- bergella in Greenland and Nova Zembla, and having examined numerous speci- mens from arctic Siberia and North America, I feel confident that L. nesophila is a distinct species. The diagnosis reads: “Of the section: Wahlbergella; perennial, dwarfy, densely caespitose with a strong main root and woody branches; leaves crowded at the base of the flowering stems, linear-lanceolate, glabrous except the margins, which are hairy from slender, glandular hairs; stems prostrate, ascending, glandular hairy; flowers solitary, nodding; calyx cylindric, glandular hairy especially along the dark coloured nerves, five-toothed, the teeth ovate, obtuse; petals emarginate, serrate along the margins, protruding through the calyx, with distinct append- ages, colour in dried specimens purplish; stamens ten, the anthers well developed; ovary rudimentary with frequently three to four styles.” With regard to the floral structure of Wahlbergella Lindman ® has described the flower of W. apetala (L.) Fr. in his interesting and highly instructive: Contri- 1 Conspectus Florae Groenl. 2 Caryophyllaceae. Medd. om Grdénland XX XVII. K¢benhavn, 1920. 3 Synopt. Flora of North America. New York, Cambridge, 1895-97, p. 224. 4 Tillagg till Spetsbergens Fanerogamflora. (Ofvers. K. Vet. Akad. Férhdlgr. Stockholm, 1869. No. 22. : 5 Fades Repert. III. 1907, p. 338. 6 Bihang. K. Svenska Vet. Akad. Hdlgr., Vol. 12. Stockholm, 1887. Arctic Plants: Morphology and Synonymy 23 B butions to the knowledge of the flowering and fecundation of Scandinavian mountain plants. This author observed on Dovre mountain two types of flowers of this species, the one being larger but less conspicuous and with the stamens shorter; the other with the petals more developed and with longer stamens. Lychnis apetala L. has a very long (about 40 cm.), persisting primary root which remains very slender, however; similar very long lateral roots proceed from near the crown of the primary.one. While in small specimens the crown of the root bears a more or less compact rosette of leaves, such leafy rosettes are in larger specimens borne on a distinct complex of subterranean stem-bases at a distance from the crown, thus representing a pseudo-rhizome; in large specimens the diameter of the rosette measured about 12 cm. _L. affinis J. Vahl shows the same structure of the vegetative organs as the preceding but the plant is smaller. A very tall species is L. Taylorae Robinson, measuring in height about 28 em. but of the same habit as the species described above; it was collected along Mackenzie bay by Rev. Isaac O. Stringer (1893). Stellaria longipes Goldie var. Edwardsii Wats. AA fA aN CO AS cae ER Figure E. 1. Oxyria digyna (L.) Hill, showing the stolons and a rosette of leaves with a flower-bearing stem; natural size; specimen from Bernard harbour. 2. Lychnis apetala L. var. arctica Th. ¥r.; flower- bearing stem; natural size; specimen from Wollaston land. 3. A petal of same; enlarged. 4. Lychnis nesophila nob.; flower-bearing stem; naturalsize; specimen from Mansfield island, Hudson bay. 5. A petal of same, enlarged. 6. Stellaria longtpes Goldie var. Edwardsii Wats.., showing a stem with leaves from the previous year, and young, floral shoots developed from the axils of the old leaves; 13 times natural size; specimen from Port Epworth. 7. Same species; part of a stolon terminated by a flower; 13 natural size; specimen from Bernard harbour. 8. Same species; a stolon; 13 times natural size; specimen from Bernard harbour. 24 8B Canadian Arctic Expedition, 1913-18 According to Warming (l.c. p. 255) “the primary root lives for a long time,” but in none of the very numerous specimens of the typical plant and the varieties which I have examined was the primary root preserved. Characteristic of this plant is the profuse development of long stolons with small scale-like leaves and stretched internodes (Fig. E). These aerial shoots are ascending and the leaves are more or less crowded on account of the shortness of the internodes. As pointed out by Kjellman? the leaves are still attached to the shoots when the winter commences but they are in a withered condition; the stems, on the other hand, remain alive and persist throughout the winter. , At the beginning of the spring small buds become visible in the axils of the faded leaves which soon develop into small leafy shoots (Fig. E: 6-8). These shoots frequently remain vegetative for one or two years until they become terminated by an inflorescence or a single flower. In specimens from Port Ep- worth the flower appeared already in the first season, as figured (Fig. E: 6). We have thus in this species of Stellaria a very interesting example of herbaceous aerial stems which winter over and produce axillary buds the function of which is to develop assimilating leaves, new axillary buds, and finally to produce flowers and fruit. This method of reproduction I observed also in the alpine plant in Colorado,? but in this the axillary shoots frequently reach the flowering stage already in the first year of their growth. — Several other species of Stellaria exhibit this type of vegetative reproduc- tion, for instance: S. longifolia Muehl., S. humifusa Rottb., and S. Holostea L. In S. crassifolia Ehrh. the withered stem-leaves subtend large buds, hibern- acula, which winter over; they have been described by Norman’, viz: ‘forma gemmificans. Caules saepe steriles vel pauciflori. Rami gemma terminali, compacta, saepius sordide violacea, ovato-globosa v. ovalia v. elliptica, usque ad 5 m.m. longa, e foliis katalyticis, carnosis, brevibus, ovalibus, obtusis, densis formata. Gemma sequente anno a ramo delapsa sensim elongatur, fibrillas radicales ramosas plures ex omni axilla protrudit, demum in caulem procumben- tem, internodiis longioribus gaudentem, se producit.”’ Merckia physodes Fisch. The primary root was not preserved in any of the specimens collected. For a comparison I have also examined material from Port Clarence, Alaska, from Moosehide mountain near Dawson, Yukon, and from Jakutsk, Siberia. Subter- ranean stolons are developed with minute scale-like leaves and with stretched internodes; capillary secondary roots develop freely from the nodi. The aerial shoots are long, prostrate, very leafy, and while many remain purely vegetative during the first season, some others develop a single, terminal flower; very often a long vegetative branch may develop from the axil of one of the leaves beneath the flower; thus it may look as if the flower were lateral but it is only apparently so. The prostraté aerial shoots develop no roots in the first year but in specimens which represented bases of old shoots roots were present, and these basal stem portions had gradually become buried in the soil. Characteristic of Merckia is thus the ability to wander not only by means of subterranean stolons but also by means of the long, prostrate, acrial shoots. Cerastium alpinum L. This forms large cushions borne upon a deep but slender primary root which branches freely. The numerous persisting stem-bases constitute a pseudo- \ ve Polarviaxternas lif, l.c. 2 Method of hibernation and vegetative reproduction in North American species of Stellaria. ‘(Am. Journ. of Science, Vol. XXV. New Haven, 1908.) 5 jilerpe Arcticae Norvegiae species et formae (Christiania Vid. Selsk. Forhdl. Christiania, 1893, p. 20. : Arctic Plants: Morphology and Synonymy 25 B rhizome and may remain active for several years. There is no leafy rosette as the cushion is formed only by the numerous branched shoots with their inter- nodes very distinct, and with the leaves opposite. Halianthus peploides (L.) Fr. This is known best under this name; it is Arenaria peploides L., but the genus has also been called Ammodenia Patrin; Honkenya Ehrh., and Adenarium Raf. The vegetative reproduction is effected by means of long subterranean stolons with membranaceous leaves, scale-like and sheathing, which traverse the ground for some time until they seek the surface where they become trans- formed into aerial shoots with green typical leaves and flowers (Fig. F: 1); many of these stolons ramify freely in the ground; thus the individual represents a mass Of floral and vegetative. shoots. Very peculiar is the development of dwarfed shoots upon the subterranean nodi of the stolons. They have been described by Warming! and, as shown in the figures (Figs. 2-4), they are either simple or ramified. The axis is very short and the leaves of minute size and scale-like; they may persist for several years without. developing any further, and, judging from the excellent material collected on the expedition, these shoots may die off without increasing any further in length. Or, as shown at the apex of the long stolon in fig. 2, they may become extended into aerial shoots. As mentioned above, these shoots may develop lateral shoots of the same structure; thus clusters of these may be seen at the root. With respect to the root-system, very thin secondary roots are developed here and there beneath the leaves of the stolons but they are not frequent and no roots were observed to proceed from the upper part of the stolons. No primary root was represented in any of the numerous specimens collected. The seedling stage has been described by Joh. Eriksson?; when growing in sand the cotyledons are hypogeic, and evidently by means of contractile toots the young plant becomes pulled down into the ground; besides true cotyledonary buds two pairs of accessory buds are developed above these. Ficure F. Haiianthus peploides (L.) Fr. 1 Part of alarge stolon bearing a lateral, more slender stolon which, on reaching the surfac) bears green leaves and a flower; natural size. 2. Part of a large stolon bearing severay dwarfed shoots, one of the apical having grown out and developed an aerial shoot; natura size; specimens from sandspit at Martin point, Alaska. 3 and 4. Two of the dwarfed shoots showing ramifications; in fig. 3 the basal leaves have dropped, leaving distinct scars; thus the shortness of the internodes are readily to be observed; enlarged. 1 Smaa biologiske og morfologiske Bidrag. (Botan. Tidsskr. 3.R. 2. Bd. Kjoebenhavn, 1877-79, p. 96, 2 Studier éfver Sandfloran i 6stra Skane. (Bih. K. Svenska Vet. Akad. Hdlgr. Vol. 22. Stockholm, 1896.) \ 26 B Canadian Arctic Expedition, 1913-18 Alsine verna Bartl. var. rubella (Wahlenb.) Lge. Some very large specimens were collected; they form cushions ot about 10 cm. in diameter, with an abundance of flowers. There is a persisting but thin primary root which is amply ramified. From the crown of the root numerous repeatedly branched shoots arise, but destitute of roots. Warming! states that in this species as well as in A. biflora, arctica, hirta, groenlandica, macrocarpa, etc. the ramification is monopodial; thus the primary shoot remains vegetative throughout the life of the plant. RANUNCULACEAE. Caltha palustris L. Although absent from Greenland, Spitzbergen and Iceland, Caltha palustris nevertheless is circumpolar; farther south it is widely distributed in Europe, and on this continent it occurs in the Atlantic States east of the Mississippi, from the mountains of Carolina and Tennessee northward to Newfoundland, thence west to Minnesota and Saskatchewan; in some forms and varieties it occurs also in Alaska, along the northern coast of Canada, and in arctic Siberia. Naturally the species exhibits a somewhat different habit throughout a range of such enormous extent, and it seems more safe to conclude that the species is polymorphic, rather than to consider the modification in habit as being of specific importance. This may be readily appreciated when we compare the species as it occurs in the temperate regions of both Worlds with the types known from the more northerly latitudes. In Norway, according to Blytt (Lc.), the stem is ascending from a decumb- ent, sometimes rooting base, and the foliage shows a reniform outline with the margin crenate; specimens with the stem rooting are by this author referred to the forma radicans Forst., known especially from Dovre mountain and Finmark. In Sweden the typical plant is common but Hartman records also the forma radicans, beside a form of diminutive size with the leaf-margin entire instead of crenate. In Russia, Ledebour describes the plant as having an erect or ascending stem with the leaves suborbicular or reniform, crenate along the margin; further- more the form with the stem rooting is also recorded by thisauthor. Ascherson ? describes the German plant as having cordate leaves at the base but reniform above; Buchenau* makes the same statement but adds that all the leaves are crenate; this author made also the observation that the variety laeta Schott, in which the follicles are erect, was the most abundant while the var. typica Huth, with the follicles recurved, appears to be very rare. In France ‘* Caltha palustris is quite frequent, and the leaves vary from reniform to somewhat cordate but always with the margin crenulate. According to Mathieu ® the basal leaves of the Belgian plant are suborbicular, reniform, and crenulate at the base while the cauline are crenulate all round. A similar leaf-shape is also characteristic of the American plant, and Torrey & describes this as “orbiculate-cordate or reniform, obtusely crenate, or nearly entire;”’ a variety integerrima (C. integerrima Pursh) is also mentioned in which the basal leaves are wholly entire but the cauline obscurely crenate. In Gray’s Synoptical Flora a variety radicans Gray is described, the leaves varying from “dilated-reniform to nearly truncate at base;” and according to this author C. radicans Forst., C’. asarifolia DC., C. arctica R. Br., and C. palustris var. sibirica Regel are merely synonyms of this variety. Finally may be mentioned that in some specimens from St. Croix Falls, Wisconsin, the margin is entire in the basal as well as in the cauline leaves, and the outline of the blade varies from reniform to completely round in the basal foliage while the stem-leaves are 1 Caryophyllaceae (l.c.), p. 241. 2 Flora der Provinz Brandenburg. Berlin, 1864, p. 17. 3 Flora der nordwestdeutschen Tiefebene. Leipzig, 1894, p. 234. 4 De Lamarck et De Candolle; Flore Frangaise. T.4. Part 2. Paris, 1815, p. 918. 5 Flore générale de Belgique. T.1. Bruxelles, 1853, p. 24. 6 Flora of New York. Vol. I. Albany, 1843, p. 17. ° Arctic Plants: Morphology and Synonymy 27 B merely reniform; on the other hand some very. tall specimens collected near ‘ Peace river landing, Athabaska, by James M. Macoun, show the leaf-margin very finely crenulate and the outline roundish reniform with the sinus relatively narrow. With regard to the arctic plant, typical Caltha palustris occurs in Nova Zembla from where it is reported by Kjellman and Lundstrém?! and where I found it myself;? my specimens, however, are very small, measuring only from 5to7cm. The forma radicans, on the other hand, is known from arctic Scandin- avia, as stated above, and Kjellman found it also on the Asiatic coast of Bering strait, besides on St. Lawrence island. Furthermore, all the specimens collected on the Canadian arctic expedition represent this form (Fig. G: 1-2); and it is known also from the islands in Bering sea. The Siberian plant, on the other hand, represents mostly the variety stbirica Regel, with the forma radicans Regel, characteristic of which is that the lobes of the leaf-blade form a very narrow sinus (Fig. 3). Figure G. 1. Caltha palustris L. forma radicans (Forst.) Hartm. A fruiting specimen from Herschel island showing three rooting shoots; two-thirds of the natural size. 2. A basal leaf of same form from the same locality ; two-thirds of the natural size. 3. A leaf of C. palustris L. var. sibirica Regel forma radicans (Forst.) Reg. from Jakutsk, Siberia; two-thirds of the natural size. In comparing the leaf-outline of these various specimens we notice that the leaf of the typical plant in Europe is mostly reniform with the margin crenulate, but varying to almost roundish, suborbicular; on the American continent the typical plant has the leaf-blade more or less reniform but, as stated above, the orbicular outline does also occur; the margin varies from crenulate to entire; the latter form is evidently C. integerrima Pursh. In Siberia the leaf-outline is generally roundish-reniform with the sinus narrow, a leaf-type that recurs in the Siberian forma radicans. In other words, the form of which the stems are decumbent and rooting occurs throughout the northern hemisphere, and the 1 Fanerogamer fran Novaja Zemlja, etc. (l.c. p. 310.) 2 Dijmphna Expedition (l.c.) 28 B Canadian Arctic Expedition, 1913-18 leaf-outline of this form agrees with that of the typical plant respectively from Europe and North America as well as from Siberia, the variety sibirica forma radicans. Caltha palustris may thus be considered a polymorphous species distributed over an area of quite considerable extent, and representing several analogous forms characterized by the structure of the foliar organs rather than by the floral. Aconitum delphinifolium DC. The tallest specimen measures a height of 25 cm. and bears several lateral racemes, shorter than the terminal one. With respect to the vegetative repro- duction there is no rhizome, and the aerial shoot is borne upon a thick but short tap-root with many lateral roots of considerable length and relatively thick. When the winter commences a bud is developed in the axil of one of the lower- most leaves, and from the base of this bud a large tap-root will develop during the winter. This manner of reproduction resembles that of A. uncinatum L. which I have described in Merck’s Report,! but differs from this by the bud remaining sessile and not being removed from the leaf-axil by means of the stolon-like stem so very plainly visible in A. uncinatum. Anemone parviflora Michx. Copious material was collected, and all the specimens represent the variety grandiflora Ulbrich.2 It belongs to the section Anemonanthea DC. of which several species have been found on this continent, for instance: A. trifolia L., known also from Europe and Siberia, A. quinquefolia L., A. oregana Gr., A. decapetala Ard. et cet. In Europe the section is more frequent; it is represented by A. nemorosa L., A. sylvestris L., A. baldensis L., A. palmata L. et cet. The section is according to De Candolle * defined as follows: ‘Car. Carpella ovoidea stylo uncinata hirsutissima villosa aut subglabra. Pedicelli in involucro solitarii aut rarissime bini umbellati, semper 1-flori nudi; sepala 5-15.” Flowering specimens are generally low, measuring only from 4 to 9 cm. in height, while fruiting specimens may reach the height of about 25 cm. The plant is perennial and has a horizontally creeping rhizome with slender stolons bearing tubular, membranaceous leaves; secondary roots are developed freely from the older part of the rhizome and beneath the aerial shoots. These shoots are erect with a subterranean ascending stem of the same structure as the rhizome, and when reaching the surface some green leaves become developed, which remain fresh during the first winter. The flower does not appear until the succeeding spring and by that time the leaves have withered. The shoots thus require two seasons to attain the flowering stage but they may persist for several years, producing alternately leaves and flowers. As stated above, all the specimens from the Canadian arctic expedition represent the variety grandiflora, and in these the diameter of the flower (dried) measured about 3 cm.; this variety occurs also in Alaska and in northern Labra- dor. The typical plant from Churchill, Hudson bay, and from Jasper park, Alberta, is a taller plant but the flower is small, the diameter not exceeding 1-7 cm. A. Richardsonii Hook. This species belongs to the same section but it is of a much more slender habit than the preceding one. The rhizome is horizontally creeping and repre- sents a monopodium until the first flowering stem develops. The internodes are stretched, slender, averaging in length from 3 to 7 cm.; one or two secondary 1 March, 1907. 2A. Engler: Bot. Jahrb. Vol. 37. Leipzig, 1906, p. 172. 3 Regni vegetabilis systema naturale. Vol. I. Paris, 1818, p. 196. Arctic Plants: Morphology and Synonymy 29 B roots, long, thin, and sparingly branched, develop close to the nodi, a little below these. With regard to the leaves of the rhizome, some scale-like leaves are developed but most of the leaves, however, are aerial, green, assimilating with the blade which is ample, deeply 5-cleft, and incised; the flower-bearing stem bears one pair of merely 3-lobed or 3-cleft, incised leaves which are sessile. Figure H. Basal leaf of Anemone Richardsonii Hook., from West Greenland; two-thirds of the natural size. There seems to be no regularity with reference to the position of the scale- like leaves; they may occur on any part of the rhizome, preceding or succeeding the green leaves. And when the rhizome ramifies the secondary branch may develop from the axil of a scale-like leaf or from the axil of a green one. While most often a green leaf is to be observed at the base of the flowering stem and subtending the bud which is to develop and continue the horizontally creeping rhizome, this green leaf may be replaced by a scale-like one. Other- wise the most frequent case is that the last leaf of the primary rhizome is aerial and green, and that it subtends the secondary branch which is to continue the growth of the rhizome in the same manner and in the same direction as if the entire rhizome were a monopodium. For instance, in a specimen from St. Paul island, Bering sea, which my late friend Mr. James M. Macoun collected for me, the rhizome measures seventy-two cm. in length, as apparently one single axis with no other sign of its sympodial structure but some few, minute scars from the withered flowering stems; on this very long rhizome only one green leaf and a flower-bearing stem were developed at the apex, and a young stolon proceeded from the axil of the green leaf. In comparing the plant as it occurs in Greenland, on the. Arctic coast of this continent, and on the coast of Alaska and adjacent islands, no particular structure of the rhizome, with reference to the foliage or flower-bearing stem, seems to be characteristic. Some more or less interesting structures may be mentioned as follows: A specimen from Beaver inlet, Unalaska, showed the main rhizome destitute of green leaves, and at the base of the flower-bearing stem a scale-like leaf subtended the stolon; but from the posterior part of the rhizome, from the axil of a scale-like leaf, a similar stolon had developed, bearing a typical, green leaf; in another specimen from the same locality the main rhizome bore two green leaves and a floral shoot; these two leaves were separated from each other by an internode about 5 cm. in length; in this same specimen the young stolon at apex bore two scale-like leaves preceding a very young, not quite developed green leaf. In a specimen from Cook inlet, Yukon river, the rhizome measured. about 20 cm. in length and bore three green leaves and a flower-bearing stem. A much shorter rhizome from Kodiak, Alaska, measuring only 7 cm., bore two green leaves and a floral stem; beside that the apical stolon bore a fully developed 30 B Canadian Arctic Expedition, 1913-18 green leaf, preceding a scale-like one and a very young green one. Finally, in some specimens which I collected in Greenland, near Holstensborg, the apical stolon had developed into a floral shoot; thus two flowering stems were developed contemporaneously. A. Drummondii Wats. Only some few specimens with the leaves withered were collected as late as ' November but the structure of these leaves agrees very well with that of this species, and much more so than with A. multifida Poir. Anemone Drummondii was also collected on the Gjéa expedition (l.c.), and the photographic repro- duction of a flowering specimen leaves no doubt as to the correctness of the identification. As far as habit is concerned, A. Drummondii is densely caespitose with a very deep, woody, primary root which branches freely towards apex. Several erect or ascending shoots arise from the crown of the root, measuring in height about 4 cm.; when reaching the surface these shoots develop rosettes of leaves and terminal flowering stems. Ranunculus Pallasii Schl. Of this stately plant several flowering specimens were collected in a pond on the tundra at Konganevik, Camden bay; they measured in length about 20 cm. We meet here with the same structure of shoot as in Anemone Richardsonit but scale-like leaves are absent. As long as the shoot is purely vegetative, the ramification is monopodial, but when the flower appears, and this terminates the axis, the shoot becomes a sympodium. Branches of R. Pallasit thus show a few stretched fistulous internodes, about 10 cm. in length, and some leaves with long petioles, subtending shoots, while opposite the leaf a tall peduncle (about 10 cm.) arises bearing a single flower; in the specimens from Konganevik some of the lateral shoots- had already developed a green leaf and a terminal flower-bud. With respect to the foliage, it may be stated that the earliest developed leaves have very small blades which are oblong or linear and borne upon long, fistulous petioles. In mature specimens the leaf-blades are larger, entire or 3-lobed, the middle-lobe being longer than the lateral ones; however, in specimens from Hudson bay (Lat. 60° 42’) the leaf-blade is 3-cleft, with the lateral lobes relatively long and narrow. The root-system consists of many very long, slender, secondary roots proceeding from the nodi. A hybrid, R. lapponicus X Pallasti, has been described and figured by Andersson and Hesselman in a paper dealing with the flora of Spitzbergen and Beeren EHiland,! previously described by A. G. Nathorst as a variety ‘‘spets- bergensis” of R. Pallasti. In this hybrid the leaf-blade is always three-cleft, the lobes entire or with the margin dentate or even lobate; the flower is smaller than in R. Pallasii and of a yellowish-white colour. A large table is given showing a number of points by which the hybrid is distinguished from the parents, and several figures of leaves, petals, stamens and fruits accompany the text. Of special interest is the fact that this hybrid is the first which has been found in Spitzbergen. R. Purshii Richards. and R. hyperboreus Rottb. were found in shallow ponds, both with the leaves floating. Of these, the former was collected with ripe fruits on Herschel island, while the specimens from Cape Bathurst were sterile, with no flowers at all, and with the leaf-blades of very diminutive size. All the specimens of R. hyperboreus were sterile with the leaf-blades very small. 1 Bihang K. Sv. Vet.-Akad. Hdlgr. Vol. 26. Stockholm, 1900. Arctic Plants: Morphology and Synonymy 31 B Ss. Figure I. 1. Ranunculus lapponicus L., specimen from Tree river, showing the apical portion of the rhizome (R.), with two green leaves (L'-L?) and a stolon (St.) developed-in the axil of L?.. One and two-thirds natural size. 2. Same species; apical portion of the rhizome (R.); S=the flower- ing stem; other letters as above; enlarged. 3. Same species; part of the rhizome with three green leaves (L'-L'); stolons are developed in the axils of L', L?, and L’; letters as above; natural size. (Figures 2 and 3 are drawn from Swedish specimens.) R. Cymbalaria Pursh var. alpina Hook. Found with the flowers fully developed, and showing the same habit as farther south, with long stolons above ground. The species is well equipped for propagation and widening of its range of occurrence, even though the individuals are very small and the flowers always single. o. R. pygmaeus Wahlenb. Numerous specimens were collected with flowers and young fruit; in spite of the small size of the plant, the shoot above ground measuring seldom more than 2 cm. in height, the vertical rhizome extends to a depth of about 2.5 cm. and bears numerous slender, ramified roots. R. lapponicus L. Like Anemone Richardsonti and R. Pallasit, this species has a horizontally creeping rhizome, and the ramification is monopodial until the first flower appears. A young rhizome (Fig. I: 1) from a specimen collected on the south coast of Coronation gulf thus shows part of the rhizome (R.) bearing two green leaves (L! and L*), while the terminal part of same rhizome (St.) is terminated by a-free bud; secondary roots, long and slender, are developed beneath the first leaf (L’). 32 B » Canadian Arctic Expedition, 1913-18 An older specimen (Fig. 2) from Norrbotten, Sweden, shows the typical structure of the mature plant. The posterior part of the main rhizome is shown by R; it bears a green leaf (L) and is terminated by a flower of which only the basal portion of the peduncle has been shown (S8.); in the axil of leaf L is a hori- zontal shoot (St.) of which the first leaf is green (L?) and of the same size and shape as leaf L'. The horizontal shoot (St.) bears at the apex a young leaf which unfolds in the succeeding season. In this way the axillary, horizontal stem or stolon (St.) will continue its growth in the same direction as the main rhizome (R.). In old specimens it thus appears as if the entire horizontally creeping rhizome were an axis of the same order but, as shown above, it actually consists of several axes and represents a sympodium. Several deviations from this structure occur, as is also the case in Anemone Richardsonii. For instance, as shown in Fig. 3, a stolon may be developed from the axil of the second leaf (L?) beside from the leaf (L') borne upon the floral peduncle (S.). But otherwise the ramification is the same because the flowering stem (S.) terminates the rhizome (R.) while the stolon (St.) is developed from the axil of the first leaf (L), bearing a green leaf (L*) which again subtends a stolon. This specimen thus shows three stolons, all being axillary. No scale- like leaves were observed in any part of these rhizomes. Ranunculus lapponicus grows in moist ground, principally in bogs, and the internodes of the rhizome may reach a length of 10 cm. or even more; the specimens from Coronation gulf were rather low, the flowering stem reaching a height of only 8 cm., while specimens from Sweden may average double the size, or more, when in fruit. R. nivalis L. and R. sulphureus Soland. exhibit the same growth and, sometimes, it has proved difficult to distinguish them from each other. To facilitate the identification a brief diagnosis of R. sulphureus given by Lange Figure k. Basal leaves of Ranunculus sulphureus Soland. (1-7), and R. nivalis L. (8-11), all two-thirds of the natural size. 1 Specimen from St. Paul island, Bering sea. 2. Specimen from Eumatowan,-Siberia. 3. Specimen from Magdalene bay, Spitzbergen. 4. Specimen from Jakutsk, Siberia. 5. Speci- men from Tromsé, Finmark. 6. Specimen from Camden bay, Alaska. 7. Specimen from Magdalene bay, Spitzbergen. 8. Specimen from St. Matthew island, Bering sea. 9. Speci- men from Dovre mountain, Norway. 10. Specimen from Cape York, West Greenland. 11. Specimen from Godhavn, Disco, Greenland. Arctic Plants: Morphology and Synonymy 33 B (Conspectus. l.c. p. 56) may be inserted here. It reads as follows: “R. altaicus Laxm. (R. sulphureus Soland.) praecedenti (R. nivalis L.) arcte affinis, a quo recedit praecipue habitu robustiore, foliis radicalibus basi rotundatis v. cuneato- contractis (nec reniformi cordatis), breviter (nec ultra medium) lobatis, caulinis fere ad basin usque 5-7- fidis, petalis pallidioribus, sulfureis, stylo breviore. In speciminibus Groenlandicis lobos fol. radic. integros, nec ut DC. (Syst. veg. 1. p. 274) indicat, dentatos invenis.”’ For comparison I insert some figures of the basal leaves of both species (Fig. K). They both are perennial, but the subterranean stem is relatively short, vertical or ascending, densely covered with old, withered leaf-sheaths; secondary roots develop from the nodes of the stem, and the primary root is of short duration, being totally absent when the plant has reached the flowering state. R. affinis R. Br. (Chloris Melvill. lc. p. 189). digitatis; lobis omnium linearibus, caule erécto 1-2-floro cum calycibus ovariisque pubescentibus, fructibus oblongo-cylindraceis, acheniis rostro recurvo. Obs. R. auricomo proxima species.” , al Lange? gives the following supplementary diagnosis: ‘Folia radicalia reniformia, leviter lobata v. magis minusve profunde palmatifida; petala pallide lutea, subtus fusco-venosa, minora et angustiora quam R. auricomi; capitulum carpellarum ovali-oblongum (nec subglobosum), ceterum R. auricomum L. affinis, sed humilior et gracilior.”’ According to Lange (l.c.) R. affinis is very rare in Greenland; it is known from East Greenland: Fr. Joseph’s fjord 73°, and from West Greenland: Arsalik in Isortok fjord. From the latter station some specimens have been figured in Flora Danica Vol. 17. Tab. 3029 (1883), and these specimens give a good repre- sentation of the species as well as the photographic reproduction of specimens from Harbour fjord in Simmons’ Flora of Ellesmere Land.? On the accompany- ing plate I have figured a flowering specimen from Bernard harbour and an achene taken from a fruiting specimen from Epworth harbour; the leaf (Fig. 2) is from Bernard harbour. While the typical species occurs in Spitzbergen, according to Nathorst,3 this author found also a plant which proved so different from typical R. affinis that he described it as a subspecies: Wilanderit. Nathorst calls attention to the following distinctive characters possessed by this subspecies: it is of low stature and caespitose; the fruiting head is roundish, not cylindrical; the carpels are thinner, with the body only a little longer than the recurved, rather coarse beak. A figure is given by Anderson and Hesselman in their interesting account of the Flora of Spitzbergen and Beeren Eiland (1.c.). J. Freyn, however, (“in lit- teris,” compare Andersson and Hesselman ].c.) on examining this plant as well as the material of typical R. affinis collected by these authors, has reached the conclusion that the specimens must be referred to R. arcticus Richards. and that R. affinis does not grow in Spitzbergen at all; furthermore, this author insists that all the material from Siberia, Davuria and Spitzbergen named R. affinis is actually R. arcticus, and that R. affinis is confined to western North America, viz.: the Rocky mountains, extending from there to Melville island. This statement seems rather peculiar when we remember that Richardson’s R. arcticus came from arctic North Ameriea, collected on the first Franklin ex- 1 Conspectus Fl. Groenl. (l.c.) p. 57. nae ; : 2 Report on the Second Norwegian Arctic Expedition in the ‘‘Fram,” 1898-1902. No.2. Christiania. 3 Nya Bidrag till Kannendomen om Spetsbergens Karlvaxter (Kgl. Sv: Vet. Akad. Hdlgr. 20. No. 6° Stockholm, 1883, p. 23. : 24657—3 34 B Canadian Arctic Expedition, 1918-18 pedition. A very detailed account of R. affinis, including R. arcticus, is given by Simmons (l.c., p. 101). However, as long as the identity of R. arcticus 1s disputed it may be well to reprint the diagnosis, inasmuch as the work in which it is published may not be readily accessible to many of the readers.'_ Richardson described the species as follows: “FR. arcticus: foliis radicalibus petiolatis hastatis tripartitis lobis divisis; caulinis in lobos lineares integerrimos partitis, caule trifolio unifloro, calyce villoso petalis breviore (B.) ? Folia glaberrima; radicalia longius petiolata, hastata, tripartita; lobo intermedio unguiculari trifido, laciniis lateralibus minoribus, patentibus; lobis lateralibus 4-partitis segmentis divari- catis, exterioribus sensim minoribus: laciniis omnibus obtusis, lanceolatis vel linearibus; caulina in lobos lineares integerrimos, obtusiusculos partita, ad bases membranacea, amplexicaulia et pilis albis mollibus ciliata: imum sub- petiolatum lobis sex, summum lobis tribus. Caulis simplicissimus pedalis erectus unifiorus, foliis tribus remotis munitus et supra folium summum pilis albis villosiusculus. Calyx flavescenti-erubescens, villosus, reflexus, petalis dimidio brevior. Petala lutea patentissima, obovata, obtusissima, receptaculo breviora. Stamina filamentis brevissimis, antheris oblongis. Germina glabra stylo recurvato mucronata, receptaculo cylindraceo.” Ranunculus arcticus Richards. was thus established on a plant which was collected on the barren grounds from lat. 64° to the Arctic sea, in lat. 69°, and it was published in the same year as Robert Brown’s Chloris Melvilliana (1823); however, in the second edition of Franklin’s Narrative Richardson adopts the name R. affinis of Robert Brown. By Gray (Synopt. Flora l.c., p. 31) R. arcticus is given as a synonym of R. affinis; another synonym is R. amoenus Karel. et Kiril. according to Ledebour (Flora Rossica, l.c., p. 732). In recent years some American authors have adopted the name R. pedatifidus Sm. instead of R. affinis R. Br. but Simmons (Flora Ellesmereland 1.c., p. 102) has clearly demon- strated that Smith’s and Robert Brown’s plant are not conspecific. Simmons, having seen the original specimens of the former, R. pedatifidus, states that the basal leaves are almost circular in outline and deeply cut into numerous, almost linear segments, with a smaller median, and two larger lateral lobes, a structure which does not occur in R. affinis; according to this author R. pedatifidus Sm. is a native of Asia, especially Eastern Siberia, besides some of the islands in Bering sea. Several very fine specimens were collected by the Canadian arctic expedi- tion, measuring in height from 6 to 12 cm. in flowering specimens; withered stems with the fruit from the year previous were about twice as tall. Part A, Plate II, Figs. 1-3 illustrates one of these specimens, and the American plant agrees exactly with that from Siberia, Nova Zembla, Spitzbergen and Green- land. The species is perennial; the primary root soon becomes replaced by a number of slender, secondary roots developed from the nodi of the short, erect rhizome. Several basal leaves surround the flowering stem which bears several long-peduncled flowers in a unilateral cyme. Ranunculus Sabinii R. Br. This is described by Robert Brown: ‘‘foliis radicalibus elongato petiolatis tripartitis: lobis ellipticis: lateralibus semibifidis; caulinis sessilibus tripartitis linearibus, calycibus hirsutis petala retusa subaequantibus. Obs. planta inter R. nivalem et pygmaeum media in Herb. D. Sabine exstat. ulterius examinanda, forsan haud distincta a R. nivali cujus cfr. icon. Flor. Dan. 1699, ubi petala retusa et folium radicale pinnatifidum.” i fren Poe Narrative of a journey to the shores of the Polar Sea in the years 1819-20-21 and 22. ondon, : 2B. denotes the Barren Grounds from Point lake to the Arctic sea. 3 Chloris Melvilliana (l.c., p. 189). Arctic Plants: Morphology and Synonymy 35 B In Gray’s Synopt. Flora (1895-97, p. 29), the species is mentioned only as a synonym of R. pygmaeus Wahlenb. ! The affinity of the species is undoubtedly with R. pygmaeus Wahlenb. and several specimens which I have examined may well be mistaken for large specimens of the latter, notably some collected by Simmons in Ellesmereland (Muskox fjord); more typical are the specimens from Cape Bathurst which are of considerably larger size with respect to foliage and flowers, and in some of these the fruiting stalk from the year previous is still preserved; it is strict and measures the height of 14 cm. The specimens from Ellesmereland, on the other hand, show the fruiting stalk arched and bent toward the ground as in R. pygmaeus. R. Sabinii is perennial, of exactly the same habit as R. nivalis and affinis as far as concerns the rhizome, the roots, and the inflorescence. PAPAVERACEAE. Papaver nudicaule L. According to Fr. Fedde! the plant from the arctic and northern part of the subarctic countries represents the subspecies P. radicatum (Rottb.) Fedde; specimens were collected at many stations visited by the expedition. The species is perennial with the primary roots persisting for several years as a deep and relatively thick tap-root which bears many slender, lateral rami- fications. Several erect or ascending shoots proceed from the crown of the root, and, on reaching the surface, they develop green leaves and flower-bearing stems, each with a single flower. Large specimens are thus of a caespitose growth in which the leaves form crowded rosettes like cushions, reaching the width of about 12 cm.; in such specimens the numerous shoots may reach the length of about 5 cm. beneath the surface; they are densely clothed with the persisting leaf-sheaths from previous years. In the tallest specimens the flowering stem measured about 14 cm., the fruiting one up to about 22 cm., but most of the other specimens were much lower. CRUCIFERAE. Lesquerella arctica (Rich.) Wats. This is a perennial herb with a deep, slender, freely branching, persisting primary root crowned with a compact rosette of leaves from the axils of which the ascending flower-bearing stems arise; the ramification of the shoot is thus monopodial. The rosette of the largest specimen measured 6 cm. in diameter; the height of the flowering stems was about 10 cm. Cochlearia groenlandica L. Collected as late as the tenth of August, at Bernard harbour, several minute specimens were commencing to bloom; in these the diameter of the rosette measured only 14 cm., the height of the flowering stem 1 cm., and the very thin, primary root about 4 cm. in length. Such small specimens were growing with large ones, about 5 cm. high, and with almost mature pods. Draba alpina L. A large number of specimens were collected of this species, the largest forming a compact cushion’ measuring in diameter 10 cm., with the flowering stems only 3 cm. in height. The primary root persists throughout the life of the plant and attains a considerable length, but is generally slender; only in one specimen did it measure about 3 cm. in thickness. In some very old specimens 1A. Engler: Das Pflanzenreich. Leipzig, 1909, p. 376. 24657—3} ‘ 36 B Canadian Arctic Expedition, 1913-18 from Bernard harbour did the height of the cushion measure 5 cm.; in such specimens the structure of the shoot could be followed from the earliest stage of the plant. From the crown of the primary root numerous shoots proceeded, densely covered with appressed, withered leaves and reaching a height of about 4 cm., whence a system of terminal and lateral rosettes commenced. In other words, the complete vegetative system of branches and leaves is above ground and winters over as such. f Among the plants which Kjellman studied at the most northerly point of Asia, Cape Tscheljuskin, Draba alpina represented the remarkable habit of forming compact balls of which the greater portion of the shoots, and especially the youngest ones, were completely above ground. A corresponding habit was also observed in Eritrichium villosum, Saxifraga serpyllifolia, Papaver nudicaule, Stellaria longipes et cet. co D. nivalis Liljebl. and D. fladnizensis Wulf. These species from Camden bay and Bernard harbour illustrate exactly the same habit as observed in ‘D. alpina. D. hirta L. D. hirta L., on the other hand, does not seem to persist for so many years as the three foregoing species. It is a tall plant as compared with most of the other Drabae, reaching the height of about 20 cm. when in bloom, such specimens having been collected on the south coast of Coronation gulf. The primary root is relatively short and thin and only a few rosettes of leaves are developed, each of which is terminated by an inflorescence. Braya purpurascens (R. Br.) Bunge. Has a monopodial shoot, the youngest specimens showing very plainly a central, leafy axis and two lateral inflorescences; in old specimens the ramifica- tion is obscured by the several leafy shoots being crowded, and the lateral position of the floral stems is only indicated by these being ascending, not strictly erect. B. glabella Richards. Only a single specimen was collected of this very.rare species, known only from the arctic coast of this continent, from East Greenland, and from a few stations in arctic Scandinavia. It is readily distinguished from B. purpurascens by the leaves being remotely dentate, and by the long, linear pods. The speci- men from Wollaston land measured a height of about 12 cm., and although being a young specimen eleven flowering stems were developed from the small rosette of leaves; the primary root is relatively short, slender, and much branched. Eutrema Edwardsii R. Br. In young specimens the primary root is slender; there is no rosette and only two to three long-petioled leaves at the base of the single, terminal, flower- bearing stem. In the older specimens the primary root is quite thick, distinctly wrinkled, and as many as six flower-bearing stems are developed from the crown of the root, besides several green leaves. The height of the flowering stem aggre- gated about 15 cm., that of the fruiting stem very nearly 20 cm. Hesperis Pallasii (Pursh) T. et Gr. [H. pygmaea (Adams) Hook.] In Gray’s Synoptical Flora it is described as a dwarf biennial plant, but, according to the numerous and exceedingly well preserved specimens collected by the expedition, it is a perennial and not always of dwarfed stature either; 1 Ur Polarvaxternas lif (1.c., p. 475). Arctic Plants: Morphology and Synonymy 37 B a flowering specimen from the south coast of Coronation gulf measures in height no less than 18 cm., and a fruiting specimen from Herschel island 10 cm., the “mature pods averaging about 8.5 cm. in length. The specimens from Bernard harbour, on the other hand, are very low, representing compact cushions of leaves from 2 to 6 cm. in diameter; in such specimens the flowers are barely raised above the leaves. As stated above, the plant is perennial which is readily to be seen from the subterranean stem-portion bearing numerous remnants of leaves from previous years and emitting short lateral branches with rosettes and flowers; moreover, in several old specimens small leafy rosettes were developed, but still too young to produce flowers. The large fruiting specimens from Herschel island have many green leaves in clusters, indicating that future growth is secured. The primary root persists throughout the life of the plant and in- creases quite considerably in thickness and length, ramifying freely, the lateral branches equalling the primary in length. Cardamine pratensis L. This seems to thrive well on the arctic shore; the specimens are tall and blooming freely; a specimen from Port Epworth harbour measures 24 cm. in height and bears three lateral racemes developed from the axils of the cauline leaves. It is a point of interest to notice that in some of these high northern specimens (Coronation gulf) vegetative reproduction is secured by means of adventitious buds developed upon the basal leaves close to the leaf-segments. These buds upon the leaves of Cardamine were described already by Cassini; while still in connection with the leaf they develop some small leaves and roots and, when liberated, they continue their growth and develop new independent individuals. Similar buds are known also from a few other plants. Turpin observed them in Ornithogalum, Henslow in Malazis, and according to Lindley ! they abound in Bryophyllum and Tellima grandiflora. With regard to the fruit in Cardamine, the specimens showed mostly flowers and only a few young pods were developed. Thus in case of failure to produce ine aaa would be able to become distributed by means of the adventi- tious buds. C. digitata Richards. In a monograph of the genus Cardamine O. E. Schultz ? rejects Richardson’s name ‘‘digitata’”’ because the genus has become merged into the genus Dentaria of which there is a species ‘‘digitata”’ of older date, thus necessitating the creation of a new name for Richardson’s plant. “Dentaria digitata Lamarck’ must consequently be renamed Cardamine digitata (Lam.) Schultz, and “C. digitata Richards.”’ must bear the name Carda-. mine hyperborea Schultz. However, it all depends on the concept of the two genera Cardamine and Dentaria, whether they should be kept separate or be united. This question has been raised before by some of the ablest botanists abroad and on this continent and the question seems to be well solved by Gray, Greene, Nuttall, Sereno Watson, and several others who were‘familiar with the species of both genera as represented on this continent, maintaining the genus Dentaria as distinct from Cardamine. With respect to the subterranean stem, C’. digitata possesses a slender, horizontally creeping rhizome which bears aerial, green leaves and which is also stoloniferous. In this way the species is better equipped to become dis- tributed than C. pratensis in case of failure to produce mature seeds in unfavour- ~ able seasons. By the constantly smaller size of the flowers, the shape of the leaves, and especially by the rhizome, C. digitata is a good species, and well distinguished from C. pratensis. 1 Introduction to Botany. London, 1832, p. 50. 2 Engler’s bot. Jahrb. Vol. 32. Leipzig, 1903, p. 550. 38 B Canadian Arctic Expedition, 1918-18 Parrya macrocarpa R. Br. [Matthiola nudicaulis (L.) Trautv., Parrya nudt- caults (L.) Regel.] Several flowering and fruiting specimens were collected; during the flowering the height of the stem, averages about 10 cm. The primary root is deep, and very thick, bearing at the crown some few short erect or ascending shoots, densely covered with remnants of old leaves, and terminated by a fascicle of leaves surrounding the flowering stem. With regard to the vegetative repro- duction, it would appear as if the species is very poorly equipped, since there is no other subterranean stem than the very short pseudo-rhizome, described above. However, the root-system shows that the plant is by no means unable to wander and thus become distributed in a vegetative manner since the large tap-root frequently branches and some of these branches, especially the hori- zontal ones, are able to produce new plants as root-shoots. . For instance, some flowering specimens collected on Herschel island and Wollaston land were actually root-shoots, with the mother-root still attached. - Similar root-shoots occur in a number of plants; Wittrock! has written an interesting paper on this subject furnishing a comprehensive list of species that multiply in this manner. Among the Cruciferae Wittrock mentions some species of Arabis, Cardamine resedifolia, Lepidium latifolium, Alliaria, and several Nasturtia. P. arctica R. Br. A relatively smaller plant than P. macrocarpa but the root is also here quite thick and deep. The aerial shoot, inflorescence, and leaves, agree with that of the preceding, but I found no specimens developed from roots. Erysimum inconspicuum (Wats.) MacMill. [#. parviflorum Nutt., non Pers.]} ‘ The tall flowering stem (about 24 cm.) arises from a small rosette of leaves borne upon a slender but persisting primary root which ramifies freely; the species is perennial. Sisymbrium sophioides Fisch. On Herschel island this species may persist for more than one season, a8 shown by a very large specimen with six flower-bearing stems reaching the height of about 30 cm. and arising from a large rosette of leaves; the thick primary root bears evidence of having been active for at least two seasons. The specimen was collected in the month of August and with it several much smaller specimens were collected. Averaging only 3 to 5 cm. in height, these plants were in bloom, and the rosette of leaves was perfectly fresh; thus they might be able to winter over. It deserves attention that this species is generally described as an annual and that it occurs as such in arctic Scandinavia (70° N.L.) according to Blytt;? it has also been found near Ivigtut, on the west coast of Greenland, but as an introduced weed only. CRASSULACEAE. Sedum Rhodiola DC. The very carefully lifted specimens form relatively large compact cushions with the numerous low, flower-bearing stems, about 5 cm. in height, borne upon fleshy subterranean branches with numerous withered stems from previous years, and terminating in a deep, relatively slender, primary root. As com- pared with Greenland specimens, those from Martin point, Alaska, are much less robust with respect to the size of leaves and inflorescence. 1 Botan. Notis. Lund, 1884. 1 Norges Flora, p. 995. Christiania, 1861. Arctic Plants: Morphology and Synonymy 39 B SAXIFRAGACEAE. Chrysosplenium alternifolium L. var. tetrandrum Lund. [C. tetrandrum Th. Fries.] This is a very small perennial herb. With respect to the morphological structure, the variety agrees with the typical plant, as described by Irmisch.! The almost filiform primary axis consists of long internodes and bears small, scale-like leaves with a minute, rudimentary blade. At the apical part of the stem the leaves are green, long-petioled, and bearing a roundish, reniform blade.” Secondary roots develop freely from the internodes. Flowering as well as purely vegetative shoots may develop from the axils of the lowermost green leaves; more frequently, however, stolons are developed from the axils of these leaves. The stolons have long internodes, and the leaves are scale-like or, towards apex, green and of the typical shape; axillary stolons develop some- times from the leaf-axils of the mother-stolon; thus the plant shows quite a compact mass of shoots, of which the stolons gradually become separated from the mother-axis and develop new, independent individuals. Parnassia Kotzébuei Cham. et Schl. The species is caespitose with a very short, erect, subterranean stem from which numerous long, slender, secondary roots are developed. There are no scale-like leaves, and the foliage forms a small rosette from the centre of which the flower-bearing stems arise. Judging from the structure of some young specimens, it appears as if the first flower-bearing stem terminates the primary shoot, and the vegetative reproduction is secured by means of one or several buds developed in the axils of the rosette leaves. These buds may be purely vegetative for a season or so and they may be separated from the mother plant and give rise to new individuals. Beside these vegetative buds there are also some others which are floral and develop during the succeeding year. No distinction was to be observed with regard to the external structure of these buds; they were not protected by scale-like leaves and their position in the rosette was not the same; thus flowers and vegetative shoots may be seen in the same specimen to have developed in no particular order. The species thus agrees with P. palustris L., according to the description given by Hj. Nilsson (Dikotyla jordst. l.c., p. 168). Saxifraga [Tourn.] L. The fifteen species collected by the expedition represent the following sections according to Engler:2 Nephrophyllum Gaud., Hirculus Tausch., Boraphila Engl., Dactylotides Tausch., Trachyphyllum Gaud., and Porphy- rion Tausch. The number of species enumerated and described by Engler (1.c.) aggregates 166, and many of these exhibit a geographical distribution of enormous extent, throughout the northern hemisphere, and to the highest eleva- tions in the mountains. Owing to this very wide distribution the genus naturally exhibits quite a num- ber of morphological structures which may well be looked upon as epharmonic characters. Nearly all the species are herbs, and they are generally small plants, but with the flowers quite conspicuous and of delicate structure; very character- istic of the genus is the immense variation in leaf-outline; we have a number of species that exhibit the same leaf-shape as various genera of other families. ‘For instance, there is a species called “aconitifolia,”’ and among others may be mentioned “ajugaefolia,”’ “cortusaefolia’”’ “hederacea,” “hieraciifolia,” ‘“heucheri- folia,”’ “juniperifolia,”’ “leucanthemifolia,” “parnassifolia,” “ranunculifolia,” etc.—As stated above, the habit is extremely different, and regardless of the 1 Zur Morphologie der monocotylischen Knollen und Zwiebelgewachse. Berlin, 1850, p. 192. 2 Monographie der Gattung Saxifraga. Breslau, 1872. , 40 B Canadian Arctic Expedition, 1913-18 nature of environment several very distinct types of habit may be observed at the same station, whether in the far north or in the alpine regions of the moun- tains. Thus among the plants which are growing nearest the Pole, between _ 82° and 83° 24’ N., gathered by the British polar expedition, 1875-76, and the late Lieut. Lockwood, United States Army, we find not less than six species of Saxifraga, viz.: 8S. cernua, S. nivalis, 8. decipiens, S. tricuspidata, S. flagellaris, and S. oppositifolia. And according to Hooker! S. cernua and S. flagellaris ascend to an elevation of above 17,000 feet in the Himalayas; furthermore, in the Rocky mountains, Colorado, I collected 8. cernua, S. flagellaris, and S. nivalis above 14,000 feet alt. In other words, the sections which are able to thrive at the stations farthest north are Nephrophyllum, Boraphila, Dacty- loides, Trachyphyllum, and Porphyrion; three of these: Nephrophyllum, Boraphila, and Trachyphyllum, are among those that ascend to the highest latitudes. . As pointed out by Engler (1.c. p. 57), the majority of the species are moun- tain plants and many of these belong to the alpine flora. This author gives an interesting table according to which the largest number of species occur in the European Alps from France to Croatia (42 sp.); then follow Tibet and the Himalayas with 35 species, the Pyrenees with 30, the Carpathian mountains with 25, the Rocky mountains with 22, etc. Only 5 species are credited to the South American Cordilleras. The majority of the species are herbaceous but very few are annual, and several may be designated as undershrubs. Among the perennial herbaceous ones some interesting structures are to be observed with respect to the vegetative reproduction, the development of runners above the ground, and of bulblets developed in the axils of the leaves, the basal as well as the cauline, in the latter case representing transformed flowers as in S. cernua and S. stellaris forma comosa, for instance. But although the primary root persists in many species, no instance is known, so far, where the root increases in thickness to any great extent, as is otherwise a feature characteristic of various mountain plants, especially the alpine ones. The arctic species collected by the expedition constitute actually an assemblage of types in which the vegetative reproduction gives an excellent illustration of the biology of the genus in the arctic and alpine regions. Saxifraga aestivalis Fisch. et Mey. ’ The plant upon which Linnaeus established the species came from Siberia, and according to Ledebour the species occurs also in Alaska and arctic America. However, some Siberian specimens which I have before me, collected near Jakutsk (N. H. Nilsson), and at Tolstoi Nos (M. Brenner) differ from the Alaskan as well as from the arctic American by the leaves (Fig. L: 3) being of a more round- ish outline and of a thinner texture, besides by the flowers being considerably smaller. The structure of the rhizome, however, is identical (Fig. 1); it is ascend- ing or sometimes vertical, quite thick, and bears numerous leaf-scars from green leaves while scale-like ones are entirely absent. Slender secondary roots are developed beneath the leaves, and the primary root evidently dies off at an early stage. The leaves form an open rosette and the flower-bearing stem (St. in fig. 1) terminates the shoot; an axillary bud remains dormant through the winter and produces a rosette of leaves and an inflorescence during the next spring. Lateral branches develop often from the rhizome, especially when growing in Sphagnum bogs. While thus the Siberian plant differs somewhat 1 Introductory Essay to the Flora Indica. London, 1855, p. 221. Arctic Plants: Morphology and Synonymy 41 B from the arctic American, the several specimens which I have collected in the Spruce-zone of the Rocky mountains in Colorado agree in all respects with the Siberian plant. Ficure L. 1. Sazifraga aestivalis Fisch. and Mey., showing the rhizome, the basal leaves, and the base of the flower-bearing stem (St.); one third of the natural size; specimen from south coast of Coro- nation gulf, Cape Barrow. 2. Same species var. Nelsoniana (Don); a basal leaf; one third of the natural size; specimen from Herschel island. 3. S. aestivalis; a basal leaf; one third of the natural size; specimen from Siberia; Jenisei, Tolstoi.Nos. S. Nelsoniana Don. This is by Engler referred to the preceding species as a variety; the leaf- blade (Fig. 2) resembles that of the Siberian plant but the petioles are much longer and the flowers relatively larger; the rhizome shows the same structure as described above. S. Lyallii Engl. and S. davurica Pall. These are both known from Alaska and adjacent islands; they exhibit the same habit as the preceding species but the leaf-shape is quite distinct. S. hieraciifolia W. et K. and S. nivalis L. These are also members of this section (Boraphtia) but their habit is very different from that of S. aestivalis. The rhizome is relatively short but quite stout in the former, and secondary roots develop freely; the almost sessile leaves are arranged in a rosette surrounding the terminal, flower-bearing stem. Axillary buds sometimes develop rosettes contemporarily with the terminal, and several inflorescences may appear upon the same individual. Characteristic of these species of Boraphila is the early withering of the primary root. Although quite frequent in the arctic region, S. stellaris L. var. comosa Poir. was not found by the expedition. I mention it in connection with the other species since it is quite an interesting plant showing some structures which recur in members of the section Nephrophyllum. The typical S. stellaris L. has an ascending or almost horizontal rhizome which, according to the substrate, may attain quite a considerable length; for instance, in specimens growing in wet moss the rhizome measures the length of about 18 cm. (specimens from Austria) and bears small, scale-like leaves subtending stolons of the same structure as the main rhizome. In these specimens the aerial leaves are so remote that no rosette is formed; however, in the northern plant a rosette i 42 8B Canadian Arctic Expedition, 1918- 18 developed from the centre of which the flower-bearing stem arises. The primary root dies off at an early state but becomes substituted by long, slender, secondary roots proceeding from the rhizome close beneath the rosette. Figure M. 1. Sazifraga stellaris L. var. comosa Poir.; natural size; specimen from West Greenland. 2. Petal of same; enlarged. 3. A bulblet of same; enlarged. 4. A bulblet of S. cernua L.; enlarged. 5. A germinating bulblet of same; enlarged. 6. S. radiata Small; natural size; specimen from Herschel island. 7. Petal of same; enlarged.” 8. S. decipiens Ehrh. from Cockburn point; natural size. 9. A leaf of same; enlarged. An arctic variety of this species is the so-called comosa (Fig. M: 1, 2, 3) in which most of the flowers are transformed into small bulblets. These bulblets represent an important means of vegetative reproduction since they are able to develop new individuals when separated from the mother-plant, thus reminding of S. cernua L. of the section Nephrophyllum. When the bulblet germinates, it produces a short, ascending rhizome, terminated by a rosette of typical leaves, and an inflorescence bearing few flowers but many bulblets. Arctic Plants: Morphology and Synonymy 43 B In the section Nephrophyllum we meet with several species which develop bulblets upon the rhizome as well as in the inflorescence; this peculiarity was known already to Linnaeus when he described S. granulata “radice granulata.”” Afterwards the morphology of this plant has been described so very excellently by Irmisch (1.c. p. 190). Among the arctic representatives of this section, S. cernua L. is the most frequent one. The rhizome is very short and the primary root must be of a very short duration since it constantly lacks in mature specimens. In this species bulblets (Figs. 4-5) are developed not only in the axils of the basal, green leaves but also in the inflorescence where they sometimes replace all the flowers. Most frequently the bracts of the inflorescence subtend only single bulblets, corresponding to single flowers, but in some remarkable specimens gathered by James M. Macoun at Fullerton, Hudson bay, the bracts subtend numerous clusters of bulblets; in these vigorous specimens the inflorescence was profusely decompound instead of being a simple raceme as in the typical plant. On the other hand, in S. debilis Engelm., which Engler considers to be merely a variety of S. cernua (1.c. p. 107), these bulblets are totally absent. The bulblets of the inflorescence are of a pink colour and they germinate freely when separated from the mother-plant. According to Warming,’ slender stolons may also occur in this species, and the foliage of such stolons may consist of scale-like as well as of green leaves some of which may subtend bulblets. S. radiata Small.3 This is S. exilis Stephan (1822) but the latter name has been rejected since there is a S. exilis Poll. of earlier date (1816). The species S. radiata, (Fig. M: 6-7) resembles S. sibirica L. very much, but in S. radiata the stem is pubescent with glandular hairs intermixed; the basal leaves are merely 5-7-lobed, and the petals are very distinctly three-nerved. Bulblets occur in both species, mainly in the axils of the basal leaves, seldom in the inflorescence. S. rivularis L. This little species forms small cushions, but the primary root is substituted by a dense mass of filiform, secondary roots; the rhizome is very short and bears rosettes of leaves with terminal flowering stems. When growing in loose, wet soil, or in moss, the rhizome may develop horizontal stolons (specimens from Camden bay), which bear typical green leaves. Bulblets occur also in this species as described and figured by Warming (l.c. p. 211), and by G. Lindmark.! The section Hirculus comprises mostly Himalayan species, but one of these, S. Hirculus L., is also widely distributed throughout the northern hemis- phere and reaches the polar regions in both Worlds. The arctic plant is low, che flowering stems reaching a height of only 6 cm. or less; it forms small and more or less compact cushions of numerous leaves, and several flower-bearing stems are produced on the same individual. There is no primary root in mature specimens but the profusely branched, subterranean shoots bear many filiform, secondary roots. The subterranean stem represents in these arctic specimens from Bernard harbour, Herschel island, etc., a horizontally creeping rhizome, about 7 cm. in length, slender, but woody, densely covered with remnants of withered leaves and branching freely; stolons with scale-like leaves are known to occur in the southern plant but I observed none in the numerous arctic speci- mens, which were examined. Vegetative reproduction is scantily represented by the arctic plant, since the branches of the rhizome are of a solid, woody structure, not permitting the secondary rosettes to become separated from the mother-plant. 1 Compare: the author in Dijmphna-Expeditionen’s Zoologisk-botan. Udbytte (l.c.). 2 Saxifragaceae (Medd. om Grgnland. K@benhavn, 1912). 3 North American Flora, Vol. 22, 1905. Jets . + Bidrag till kannedomen om de Svenska Saxifraga-Arters yttre byggnad ach individbildning. Bih. K. Sv. Vet. Akad. Hdlgr. Vol. 28. Stockholm, 1902. 44 B Canadian Arctic Expedition, 1913-18 S. decipiens Ehrh., incl. var. groenlandica (L.) Lge., and S. silenaeflora Sternb. These are the only members of the section Dactyloides which are known from arotic America; of these only the former was collected. S. decipiens Ehrh. is the name adopted by Engler in his Monograph of the genus (1.c. p. 186) and the one used by the various authors, who have written on arctic botany, for instance: Kjellman, Trautvetter, Lange, and others; recently, however, another name has been introduced, viz.: S. groenlandica L. by H. G. Simmons." S. decipiens is densely caespitose, and the primary root persists; it is slender but deep and ramifies freely; secondary roots, on the other hand, develop seldom. There is no rhizome in the proper sense of the word, but numerous shoots develop from the crown of the root. Lindmark (1.c.) mentions that lateral shoots appear at an early stage and that they are developed from the axils of the lowermost leaves of the main shoot. The accompanying figure (Fig. M: 8) shows only a small part of a large cushion; the shoots are densely clothed with withered leaves, and at the apex they bear several rosettes of fresh leaves surrounding the flower- bearing stems; since the development of secondary roots is very scant, the shoots do not, so far as I know, become separated from the mother-plant; thus the species is evidently not able to multiply by means of vegetative reproduction. While Engler (1.c.) has described a number of varieties of the species, Lange (l.c.) mentions only those occurring in Greenland, as follows: 1. groenlandica “2-4-pollicaris, dense pulvinata, glanduloso-pilosa, foliis trifidis, laciniis obtusis subintegris;” 2. uniflora: “humilis, 1-3-flora, calyce nigro-glanduloso, ceterum ut 1, in quam sensim transire videtur, nam formae uni- et pluriflorae promiscue leguntur;” 38. Sternbergii: “spithamaea v. ultra, laxius caespitosa, minus dense glanduloso-pilosa; foliis palmato-trifidis, laciniis lateralibus bi-trilobis.”’ The section Trachyphyllum comprises types of very distinct habit; among the arctic we meet with S. bronchialis L. and S. tricuspidata Rottb. which have the same habit as S. decipiens; S. atzoides L. which in several respects reminds of S. H treulus, and finally the peculiar S. flagellaris Willd. Of these, S. tricuspidata Rottb. forms large, compact cushions in the same mahner as S. decipiens, and the primary root persists, while secondary roots were not observed. In this species the flower-bearing stem is quite tall, reaching a height of about 10 cm. A specimen from Greenland, Pagtorfik, Noursoak peninsula, differs from the type by the shoots being very long, about 23 cm., and the leaves remote, thus forming no rosettes. S. bronchialis L. agrees with the former but besides the persisting primary root secondary roots are also present, developing freely from the subterranean stems. An interesting variety is cherlerioides Don, in which the lateral shoots are extremely short with the minute leaves crowded and thicker than in the type; itisanative of eastern Asia but has also been found in Alaska and adjacent islands. Moreover, the species occurs also in the Rocky mountains, ascending about 12,000 feet above sea-level in Colorado (Gray’s peak), and specimens from this locality as well as from lower altitudes, for instance in Clear Creek canyon (10,000 feet), agree well with the Siberian plant; but when compared with the Alaskan plant (Kodiak, legit Walter H. Evans) there is some difference with respect to the fruit; this is considerably larger in the Alaskan specimens but otherwise structure of the leaves and their position, being densely crowded, is identical. Some years ago Dr. Wiegand established a new species, S. austromontana, founded upon specimens of S. bronchialis from the Rocky mountains, with the following distinction: ‘‘Differs from S. bronchialis L. in its more subulate, darker green leaves, with fewer ciliae near the apex; more slender pedicels, smaller, 1 Flora of Ellesmereland (l.e. p. 70). Arctic Plants: Morphology and Synonymy 45 B white petals with purple dots above, and not unguiculate at the base; stamens shorter than the corolla; and smaller capsule with much shorter styles. It probably includes all of the so-called S. bronchialis from the Rocky mountains south of Alaska.”’ Some: specimens which I have before me, collected by Mr. Crandall on Pennock’s mountain ranch in Colorado (altitude 7,500 feet) agree exactly with the distinction pointed out by Dr. Wiegand, but I do not believe these dis- tinctive characters are sufficient for the establishment of a new species. Professor Engler in his excellent monograph of the genus (l.c. p. 215) calls attention to the variability of S. bronchialis L. corresponding with that of ‘S. aspera DC., of which the alpine form is more densely caespitose, and with the inflorescence few-flowered in contrast with the plant from lower elevations. Finally, if we compare S. flagellaris Willd. as represented in the north and south, we notice quite a prominent difference in the relative size of flower and fruit, the structure of the foliage,-etc., but without feeling inclined to divide the species. From the experience I have had with Sazifraga, especially in the field, I should prefer to consider S. austromontana Wieg. as a form of S. bronchialis L., instead of an independent species. S. aizoides L. Several complete specimens were collected by the expedition, “complete,”’ because in several of the specimens the primary root was preserved. However, the arctic plant differs from the southern by its more condensed growth; quite compact cushions may be developed, measuring in width about 7 cm., not including the flowering stems. Among those from Bernard harbour some specimens are of quite diminutive size, barely 2 cm. high, and it is in such speci- mens that the primary root was present; it is relatively deep, but slender, and emits several branches; secondary roots develop freely from the subterranean stems. In the southern plant the stems are more spreading, in the manner of S. Hirculus, but typical stolons with scale-like leaves were not observed. In the arctic plant the leaves are much crowded and the shoot is quite profusely ramified; thus a cushion may be formed. Vegetative reproduction is secured by means of the lateral shoots, which may be separated from the mother-plant and, by their ability to develop secondary roots, new individuals may thus be formed. S. flagellaris Willd. This is one of the most interesting species of the genus. We find it in the polar regions as well as in the alpine, and exhibiting principally the same morpho- . logical structure. The arctic plant is generally of very low stature, the flower- bearing stem reaching a height of only 13 to 3 cm. The shoot bears a number of fleshy leaves, glandular hairy especially along the margins, and these leaves form a small rosette resembling that of a Sempervivum. In spite of the very considerable material which I have examined, there is not a single specimen showing a persistent primary root; the root-system consists only of a number of secondary roots, developed from the basal internodes of the rosette. A single flower-bearing stem, bearing a few leaves and one or two flowers, terminates the shoot. Very characteristic of the species is the development of long runners above the ground, consisting of a single internode, reaching a length of about 10 cm., and terminated by a small spherical rosette of green leaves; these runners are developed in the axils of the rosette-leaves. When the fruit is mature the main shoot dies off and at this time, towards the end of the season, the rosettes borne on the runners have commenced to develop roots; as the runners die off with the mother-plant, these rosettes thus become independent individuals and repeat the growth of the mother-shoot. I do not know how long time these rosettes require to become mature to produce flowers, but judging from the size of various specimens which I collected in Nova Zembla, I presume they reach maturity in the third year. 1 Bull. Torrey Bot. Club. Vol. 27, 1900, p. 388. 46 B Canadian Arctic Expedition, 1913-18 In comparing the arctic with the alpine plant, for instance with a series of specimens from the Rocky mountains of Colorado, the following deviations were noticeable. The flowering stem is taller, more leafy, and two to three flowers may be developed; moreover the plant is more glandular hairy. Finally, in some specimens from near the summit of James’ peak (13,000 feet), a somewhat remarkable structure occurs.. They grew in somewhat damp soil associated with Sieversia, Dryas, Actinella, etc., and the flower-bearing stem reached a height of about 15 em., bearing about 7 flowers in a unilateral cyme; the flower- bearing stem was very leafy and several of the basal leaves above the rosette subtended runners of the usual structure. The rosette was not so compact as 1n typical specimens and, moreover, a subterranean stem-portion, about 5 cm. in length, extended from the rosette to a cluster of secondary roots; this stem- portion bore some remnants of withered leaves and consisted thus of more than a single internode. Some isolated young rosettes which grew near the flowering specimens showed a similar elongated stem beneath the rosette-leaves, provided with a corresponding system of secondary roots at the lower end of the stem. A third type of‘specimens, however, explained this singular structure; it con- sisted of a rosette of leaves with runners, but instead of being terminated by an inflorescence the main shoot had continued to grow above the rosette as a vegetative shoot bearing several scattered leaves and terminated by a rosette of a more open structure than in the typical plant. In other words, the alpine S. flagellaris may remain at a purely vegetative stage for several years, but not as a single rosette, gradually increasing in size, as is the case in the arctic speci- mens, but developing an erect, purely vegetative shoot, of which the apex assumes the shape of a rosette to produce flowers in the succeeding year, and still depending on the same fascicle of secondary roots. The age of such specimens appeared to be not less than four years. The fact that none of the specimens examined possessed a primary root naturally indicates that they owed their existence to rosettes of runners, which undoubtedly is the most common method of reproduction in this species. However, capsules with ripe seeds are frequently to be found in alpine specimens, and even in Nova Zembla did I succeed in finding fully matured capsules with seeds. The section Porphyrion-to which S. oppositifolia L. belongs: comprises only three other species: S. biflora All., 8S. macropetala Kern., and S. retusa Gouan. Of these the two last are natives of the mountains of Middle and South Europe; S. biflora is also a native of these mountains, but extends northward to Lapland and northern Russia. S. oppositifolia, on the other hand, is widely distributed throughout the northern hemisphere and abounds in the polar regions. It is actually an under- shrub, since the profusely ramified branches above ground are woody. The long stems are mostly prostrate with the internodes more or less stretched, with the leaves opposite, and developing secondary roots quite freely. Erect, but very short, floral shoots are developed in the axils of the cauline leaves, and they bear only one flower, at first almost sessile, but distinctly pedicelled when fruiting. In very large, old specimens the posterior parts of the stem have frequently died off; thus the root-system is confined to secondary roots; but in smaller, younger specimens, the primary root may be preserved; it is slender, but quite deep, and bears several lateral branches. : An interesting variety, Nathorsti, was found by Dusén in East Greenland 4 and described as follows: ‘‘Laxe caespitosa; ramuli steriles breves, c. 3 cm. longi, sat dense foliati, foliis oppositis; ramuli floriferi elongati, 5-6 cm. longi, inferne glabri sursum sensim albopilosi, apicem versus dense pilosi, remote foliati, foliis medianis et supremis saltem alternis, rarissime oppositis; folia longiora et remotius ciliata quam in typo, rotundate acutata, usque ad 10 mm. longa et 3 1Bih. K. Sv. Vet. Akad. Hdlgr.° Vol. 27. Stockholm, 1902, p. 35. Arctic Plants: Morphology and Synonymy 47 B mm. lata; flores solitarii vel rarissime bini, majores quam in typo, colore vari- abiles, pallide rubro-violacei, pallide rosei vel albo-rosei; sepala fere triangularia, remote ciliata; petala ovalia vel oblongo-ovalia.”’ The diagnosis is accompanied by several figures showing the distinguishing characters very plainly. Gunnar Andersson and Henrik Hesselman have described two other types,! viz.: forma reptans and forma pulvinata. Of these, the former is appressed to the ground, forming loose mats, but of great width, until about 1 m. in diameter. In forma pulvinata the growth is very compact, and the width aggregates only to 20-30 cm. Dryas L. While D. octopetala L. and D. Drummondti Hook. have always been recog- nized as distinct species there has sometimes been expressed doubt about the specific validity of D. integrifolia M. Vahl. Kjellman,? who had the rare oppor- tunity to observe D. octopetala and D. integrifolia at the same station on the Asiatic coast of Bering strait at Konyam bay, felt convinced that they represent two distinct species. Nathorst,? who also found them growing together, at Cape York on the northwest coast of Greenland, observed moreover a form intermediate between both which he named D. integrifolia forma intermedia; in this form the leaves varied from being dentate only at the base to dentate along the whole margin from base to apex. This author felt inclined to con- sider this intermediate form as a hybrid, inasmuch as the two species keep remarkably constant, even when distributed over large areas. The forma intermedia has since also been reported from Northeast Greenland, viz.: Scoresby sound by Hartz,‘ who has figured an interesting series of leaves of typical D. octopetala, of the forma intermedia, and of D. integrifolia. The specimens from Port Epworth harbour represent the forma intermedia, but in these specimens the dentate leaves are very few in proportion to the entire ones, and the specimens are better referable to D. integrifolia than to the other. Sieversia glacialis R. Br. A few specimens in fruit were collected of this very rare species, the stem reaching the height of fifteen cm.; by the styles being pilose the species is readily distinguished from 8. Rossii R. Br. Potentilla palustris (L.) Scop. Two vegetative shoots were collected with relatively large leaves, the length of the leaflets being about 2 cm. and the width about 1 cm. P. fruticosa L. The specimen from Port Epworth, collected in the month of July, is quite a little shrub 17 cm. high, with several stout branches, with its leaves distinctly petiolate, and with the floral peduncles attaining a length of about 3 cm.; the flowers are large and as well as the leaves larger than those I have seen in the alpine region of the Rocky mountains in Colorado. 1 Spetsbergens Flora, l.c. p. 25. : 2 Asiatiska Beringsundskustens Fanerogamflora. Vega Expedit. vetensk. arbet. Stockholm, . 527. Wek Bensidia anteckningar frin nordvestra Grénland. Ofvers. Kgl. Svenska Vet. Akad. Férhdlgr. Im, 1884, p. 24. : ; BT urseeee og Karkryptogamer fra Nord@st-Gr¢nland. Medd. om Grdgnland, XVIII. Kjgben- havn, 1895, p. 320. 48 B Canadian Arctic Expedition, 1913-18 PAPILIONACEAE. Lupinus nootkatensis Donn var. Kjellmanii Ostf. The species is perennial with a deep, strong primary tap-root bearing a large rosette of leaves surrounding a tall flower-bearing stem, the tallest specimen measuring about 24 cm. in height, the flowering raceme itself about 8 cm. In ee specimens the large:-legumes are densely villous, and about 5 cm. in ength. Astragalus [Tourn.] L. The two species of Astragalus show a very different habit. In A. alpinus the primary root persists but it is very slender; the aerial shoots are quite long but very slender, prostrate, and profusely branched, bearing several leaves and a terminal, long-pedunculate inflorescence. Some few secondary roots develop from the nodes, and tubercles were found on some of the lateral roots. : In A. aboriginorwm, on the other hand, there is a very strong and thick, primary root, often branched and crowned with numerous, compact clusters of persisting internodes with scale-like leaves, supporting leafy shoots and flower- bearing stems; root-shoots abound. ; As may be seen from the figure (Part A, Plate X), the arctic specimen of A. aboriginorum differs in several respects from the prairie plant described and figured by Hooker (Fl. bor. Am.), but Mr. Marcus Jones, to whom specimens were sent for identification, kindly informed me that the arctic material repre- sented the typical plant. , With regard to the systematic position of this species Hooker (].c.) states that the perfect fruit which was unknown to Dr. Richardson has now been seen in abundance, and proves the species to be a Phaca, rather than an Astragalus. Phaca frigida L. In specimens from Bernard harbour, collected in the month of August, the racemes with fruit, not fully matured, measure 15 cm. including the peduncle, while in bloom the stem above ground measures only about 8 cm. in height. The primary root is long and slender and the capillary, lateral roots bear tuber- cles. As in the plant from Nova Zembla, which I have described and figured (Dijmphna Expedition 1.c.), the aerial shoots develop from the axils of scale-like leaves borne upon persisting basal internodes; thus a pseudo-rhizome is also possessed by this plant. Oxytropis Roaldi Ostt. Having been published in Christiania,! and perhaps not being readily accessible to the reader, I herewith insert the diagnosis as written by Mr. Osten- feld; it reads as follows: “Scapigera, usque 20 cm. alta, multiceps; stipulae alte petiolares, albae, membranaceae, longe ciliatae, pars libera uninervis, e basi dilatata linearis; foliola 7-8-juga, lancelato-ovata, adpresso niveo-pilosa; scapi foliis longiores pilis subadpressis vel subpatulis instructi; inflorescentia subcapitata, 5-10- flora; bracteae lineari-lanceolatae, calycem subaequilongae; calyx tubuloso- campanulata, dentibus triangularibus tubo triplo brevioribus, pube nigra pilisque longioribus albis instructa; corolla calyci dimidio longior, violaceo- purpurea; legumen (altum) ovato-oblongum, recte, acuminatum, membranaceo- chartaceum, ca. 12 mm. longum, uniloculare vel partim semi-loculare, pube breve griseo, calycem subduplo superans. Herschel Island.” 1 Vascular plants collected in arctic North America (King William Land, King Point and Herschel Island) by Gjéa Expedition 1904-1906. Vidensk. Selsk. Skrift. Christiania, 1910. Arctic Plants: Morphology and Synonymy 49 B Characteristic of the species of Oxytropis collected on the expedition is the caespitose growth with numerous leaves forming compact cushions, especially in O. nigrescens and O. arctobia The primary root. persists; it.is quite long and relatively thick. frequently supplemented by secondary roots of almost the same length and thickness. Numerous aerial shoots with the very short inter- nodes completely hidden by the crowded, withered leaf-sheaths, stipules, rises from the crown of the root, and these shoots are terminated by rosettes of, [eave surrounding the central flowering stem overtopping the leaves, as in O. foliolosa, O. campestris and O. Roaldi; or the peduncle of the inflorescence.is so short that only the flowers themselves are raised, and only slightly so, above the foliage, as in O. nigrescens and O. arctobia. The habit of these two species is thus identical with that of Trifolium nanum from the alpine region of the Rocky mountains. A specimen of O. nigrescens from Camden bay consists of a compact cushion measuring 18 cm. in diameter; the largest of O. arctobia measured 12 cm. in diameter. Common to these two species of Oxytropis and the Trifolium men- tioned above is the two-flowered inflorescence, while in the other species there are several flowers forming a head or.a short, more open raceme. Hedysarum Mackenzii Rich. shows the same habit as Oxytropis campestris, but the flowering stems are much taller and branched, specimens from Ber- nard harbour measuring 28 cm. in height; the inflorescence is racemose. Characteristic of these arctic Papilionaceae is thus the persisting primary root; in most of these the growth is caespitose owing to the profuse development of aerial shoots of which the basal internodes persist and from the leaves of which new shoots arise, bearing leaves and flowers, and dying down to the ground at the end of the season, leaving only the lowermost part to persist and repeat the ramification in the manner of a pseudo-rhizome. 7 EMPETRACEAE. Empetrum nigrum L. Several specimens were collected on tundra, with flowers and fruit; the. branches showed the typical growth of the species and were quite long, reaching 30 cm. in length. a ONAGRARIEAE. Epilobium latifolium L. The tallest flowering specimen is from Cape Barrow; it measures 20 cm. in height, and several stems are developed from the woody, thick, primary root; as to the size of the flowers, the dried petals measure 2.2 cm. in length and about 1 cm. in width. The leaves are lanceolate, quite long, and relatively narrow. In specimens from Wollaston land the flowering stems are much shorter, only 9 cm. in length, and the leaves are short but very broad, oval in outline. While thus the primary root persists and develops as a strong vertical root, the basal subterranean stem-portions persist also and increase in thickness; upon these buds become developed which give rise to the aerial, floral shoots; thus we have in this plant a pseudo-rhizome with the primary root persisting and increasing quite considerably in thickness. Among the numerous specimens collected there was no indication of reproduction by means of root-shoots. E. angustifolium L. The Fire-weed is so widely distributed in the arctic regions that it is even circumpolar, but it is nowhere as abundant in the polar regions as farther south, throughout the entire northern hemisphere. And farther south it is especially 24657—4 50 B Canadian Arctic Expedition, 1913-18 abundant in clearings and newly burned lands. It is also called Willow-herb on account of the seeds bearing a tuft of long hairs through the help of which the plant is readily disseminated over areas of great extent. However, the species possesses also another means by which its dispersal is effected, namely, by the long horizontal roots developing root-shoots; these roots with buds may persist for many years in the soil in a dormant state waiting for the oppor- tunity favourable to their farther development. And this opportunity is the cutting down or burning of the forest. This is a-fact so well known that it is hardly worth mentioning but, among the many reports that have been given about the sudden and abundant appearance of the species, one might be cited here which is quite interesting. Mr. I. W. Chickering | writes: “In northwestern Maine, on a tract of land of some 4,000 acres, over which lumbering operations had been carried on some years ago, leaving a tangled mass of limbs and under- brush a fire broke out June 8th and swept over this entire tract, lasting for two weeks, burning with such fury that it was almost impossible for the stage to travel along the road. A new vegetation began to start in three weeks after the fire, and the whole region of 4,000 acres was covered with Epilobium angusti- folium as far-as the eye could reach, over hill and valley, ridge and interval was one mass of colour from the Fire-weed.”’ This sudden and so abundant appear- ance of the plant must necessarily have been caused by the rapid development of root-shoots, which thus were in the soil in the state of buds waiting for the opportunity. From the writings of Irmisch ? we have learned about this plant that it often reaches the flowering stage in the first year. But, of course, .the vegetation mentioned by Mr. Chickering could not possibly owe its appearance. to seeds, three weeks after the fire. But Irmisch describes also the root-system, and this author made the interesting observation that the primary as well as the secondary roots of the young seedling develop buds freely which sometimes give rise to new plants in the succeeding year. The wide distribution in the north of the species evidently depends ‘on its dissemination by the wind; its persistence, on the other hand, it owes to the development of root-shoots. UMBELLIFERAE. Bupleurum americanum C. et R. The only arctic representative of the genus and in no particular morpho- logically distinct from its numerous, more southern, congeners among the perennial forms. The primary root persists as a relatively slender tap-root crowned with a rosette of long, narrow leaves and a few or only a single flower- bearing stem, somewhat taller than the leaves; the tallest specimen measured about 14 cm. in height. Selinum cnidiifolium Turcz. A few flowering specimens of this very conspicuous plant were collected on Herschel island; it shows the same habit as Pachypleurum, but is more robust, the stem taller, about 20 cm., and the ample, basal leaves deeply cut; the largest umbel measured 10 cm. in diameter.’ ERICACEAE. Ledum palustre L. In all the specimens collected the leaves are remarkably short, sometimes barely 1 cm. in length, and very narrow; the inflorescence is also much smaller than in the typical plant. Some of these specimens may be referable to the variety decumbens Ait., but in others the stems appear to have been eréct, and attaining a height of about 23 cm. 1 Bot. Gazette, Vol. 9, p. 193. Chicago, 1884. 41 Botan. Zeitung. Leipzig, 1857, p. 459. Arctic Plants: Morphology and Synonymy 51 B Rhododendion lapponicum (L.) Wahlenb. Some of the specimens are quite bushy but with the stems more or less prostrate; as to the foliage and the flowers these agree in all respects with those of the Greenland plant. Kalmia polifolia Wang. Seems to be very rare as only two small specimens were collected; the aerial branches are erect, about 6 cm. high, and are mostly one-flowered; the leaves are short and very narrow, and the specimens are even more reduced in size oe cue which I have observed in the high mountains of Colorado above timber line. Loiseleuria procumbens (L.) Desv. Only two specimens were found on the south coast of Coronation gulf, in Epworth harbour. The denuded, twisted stem is quite thick and prostrate, bearing at the apex many ascending branches with leaves and flowers of the typical structure. : Cassiope tetragona (L.) Don. Common, and collected at several stations; it varies much in height; from 8 to 22 cm.; the smallest specimens are from Camden bay; in these the prostrate stem bears numerous ascending, crowded shoots with many flowers. Otherwise the plants resemble those from Greenland. Arctostaphylos alpina (L.) Spreng. All the specimens collected show the characteristic habit of the species; in fruit-bearing specimens from Bathurst inlet (August 25th) the leaves are of normal size, measuring about 3 cm. in length and 1.2 cm. in width. VACCINIACEAE, Vaccinium uliginosum L. While all the specimens represent the variety microphyllum Lge., those from Bernard harbour are the most peculiar, the length of the leaves averaging only about 5 mm. V. Vitis-idaea L. var. pumilum Hornem. The very small specimens agree exactly with the plant as it is developed in other arctic countries; although collected as late as in August, the specimens were all in bloom. . PRIMULACEAE. Androsace Chamaejasme Host. The habit of this interesting little plant is the same as that of specimens from the alpine region of the Rocky mountains. It actually represents an undershrub with two types of shoots. The persisting, but very slender, primary root is crowned with a dense rosette of leaves from the axils of some of which long slender stolons develop. These stolons are completely above ground and consist of one or several stretched internodes the foliage of which is in the shape of rosettes, but only the apical develops an inflorescence, a scape with a few- flowered umbel, or it remains purely vegetative, giving rise to a secondary. 24657—43 52 B Canadian Arctic Expedition, 1913-18 ra SSBSS Gas WES i. OES \ f \ j A bien y ap SOWA yy SVAN Ni Si } : H } Sf Wyn hed py j x iS “UZ Ficure N. Androsace Chamaejasme Host.; a flowering specimen, showing the primary root and th® stolons above ground terminated by rosettes of leaves and inflorescences; natural size; specimen prom Bernard harbour. group of sessile rosettes. The rosette dies off after the first flowering, becoming replaced by the secondary rosettes. The root-system is thus poorly represented and secondary roots are very scarce; when such develop, they proceed, one or two together, from the nodi of the stolons. There are thus apparently two types of shoots in this plant, viz.: the stolons with stretched internodes, and the axis of the rosette, consisting of extremely short internodes with crowded leaves. However, as a matter of fact, it is the same shoot, an axis of the same order, which thus becomes modified in structure according to its function, to develop in the manner of a horizontally creeping stolon, with the uppermost internodes becoming shortened so as to produce a rosette with a terminal inflorescence. Primula borealis Duby. Kjellman,! who has offered so many and most excellent contributions to the knowledge of the life-history of arctic plants, describes the over-wintering buds of Primula nivalis. In this species the over-wintering bud consists of several fleshy, scale-like leaves surrounding a series of very small leaves which in the following season will develop into green assimilating leaves; in the centre of the bud an inflores- cence is already formed, and this structure is to be observed at the commence- ment of the winter. The material of P. borealis Duby collected on the expedition contains numerous specimens with an old withered scape from the year previous, 1 Ur Polarvaxternas lif. Stockholm, 1884. Arctic Plants: Morphology and Synonymy 53 B as well as a scape of the year of collection, bearing an umbel of flowers. In the axil of one of the innermost leaves of the rosette the bud to winter over is already visible; thus these two arctic species show the same method of hibernation. Furthermore, in both species the rhizome is vertical but extremely short, and there is no trace of the primary root; the root system consists of several long secant roots developed from the basal internodes of the subterranean stem or rhizome. Dodecatheon frigidum Cham. et Schl. _ _As in the Primulae described above, there is an over-wintering bud situated in the axil of one of the innermost leaves of the rosette, and of the same structure. But Dodecatheon possesses a very distinct, well developed rhizome which is horizontally creeping, quite thick, and reaching a length of up to 4.5 cm. Numerous long, fleshy, sparingly branched, secondary roots develop from the lower face of the rhizome. Douglasia Lindl. According to Bentham and Hooker only four species are known, three being natives of this continent, the fourth of the European Alps, generally called Aretia vitaliana Gaud. The genus belongs to the section Primuleae, and its nearest ally is Androsace. Gray (Synopt. Fl. l.c.), describes four species from this country: D. nivalis Lindl., D. arctica Hook., D. laevigata Gr., and D. montana Gr. Of these, D. nivalis is known only from near the sources of the Columbia at an elevation of 12,000 feet (Douglas), while D. arctica has been found on the arctic seashore between the Mackenzie and Coppermine rivers (Richardson), and Red mountain, Yukon valley (M. W. Gorman, 1899); since then it has been’ reported from King point, on the Arctic coast, Lat. N. 69° 6’, Long. W. 137° 40’, by the Gjéa expedition, and from between Herschel island and the Mackenzie river delta by I. O. Stringer. D. arctica forms small compact cushions of erect or ascending woody stems, densely covered with the appressed linear leaves and terminated by the flowers which are arranged in a small umbel. None of the specimens examined had the root-system preserved. GENTIANACEAE. Gentiana arctophila Griseb. and G. propinqua Richards. They both are annuals and of exactly the same habit, viz.: a small rosette of opposite leaves, a central, terminal, flower-bearing stem with one or two pairs of opposite leaves subtending' single flowers, and furthermore, one or two pairs of one-flowered peduncles much shorter than the main stem, developing from the axils of the innermost pairs of leaves of the rosette. They resemble each other very much but may, nevertheless, be readily distinguished by means of the structure of the corolla which, according to Grisebach,! shows the follow- ing characters: ‘“Corollae 4, rarius 5-fidae coeruleae lobis ovato-lanceolatis acuminatis cuspidatis setula coronatis et margine hinc setulosis tubo obconico basi angusto fere 2-plo brevioribus in G. propinqua,”—while “corollae 4 fidae coeruleae lobis triangulari-ovatis basi superincumbentibus acutis cuspidatis setula coronatis margine nudis tubo sensim ampliato basi lata 3-plo brevioribus in G. arctophila.” . Most of the specimens are very small; the height of G. arctophila averages from 6 to 7 cm., and from only 2 to 6 cm. in the other. 1 Genera et species Gentianearum. Stuttgart, 1839. 54 B Canadian Arctic Expedition, 1913-18 Pleurogyne Esch. While P. rotata Griseb. has been recorded from quite a number of stations on this continent, viz.: Labrador and Hudson bay to the high-northwest coast, Kotzebue sound, ete., and the Rocky mountains south to lat. 39°, P. carinthiaca Griseb. has not with certainty been observed here; Gray and Macoun have credited a variety pusilla Gray to Labrador and the alpine region of the White mountains of New Hampshire (Pursh), but according to Robinson and Fernald,? the American P. rotata includes P. carinthiaca var. pusilla, which by Macoun * has been reported from Anticosti and some few stations in Quebec. Meanwhile both species were collected by the Canadian arctic expedition, and since P. carinthiaca has been confused with the other, some brief notes on their distinctive characters may be appropriate. Both were described by Grisebach,’ viz.: “P_ rotata, foliis lanceolatis linearibusque, sepalis lanceolato-linearibus corollam aequantibus, ovario acuto, seminibus suturalibus,” and “P. carinthiaca, foliis ovalibus oblongisque, sepalis ovatis corolla duplo brevioribus, ovario obtuso, seminibus prope suturam insertis.” A specimen from Bernard harbour of P. carinthiaca has been drawn on Plate XI, fig. 3 (Can. Arct. Exp., Vol. 5, pt. A), 14 Figure O. 1. -Pleurogyne carinthiaca Griseb.; a flower; one and two-thirds times natural size; specimen from Almagell,- Valais, Switzerland. 2. Stem-leaf of same specimen; one and two-thirds times natural size. 3. Same species; specimen from mouth of Mackenzie river; natural size. 4. Same species; a corolla-lobe; one and two-thirds times natural size; specimen from Bernard harbour. 5. Calyx-lobe of same specimen; one and two-thirds times natural size. 6. Pleu- rogyne rotata Griseb.; a flower; one and two-thirds times natural size; specimen from Bath- urst inlet. 7. Stem-leaf of same specimen; one and two-thirds times natural size. 8. Same species; an open flower; specimen from Blagowjestschensk, Amur; one and two-thirds times natural size. 9. A closed flower of same specimen; one and two-thirds times natural size. 10. Stem-leaf of same specimen; natural size. 11. Same species; a flower; specimen from near Twin lakes, Leadville, Colorado; altitude 9,265 feet; one and two-thirds times natural size. 12. Same species; a flower; one and two-thirds times natural size; specimen from Churchill, Hudson bay. 18. Same species; a flower; one and two-thirds times natural size; specimemfrom the same locality. 14. A stem-leaf of same specimen; natural size. 1 Gray’s New Manual of Botany. 1908, p. 659. 2 Catalogue of Canadian Plants, Part II. Montreal, 1884, p. 325. 3 Genera et species Gentianearum;. Stuttgart, 1839, p. 309. Arctic Plants: Morphology and Synonymy 55 B and another specimen of the same species from the Mackenzie river delta is figured in the ‘accompanying text-figure (Fig. O) beside a specimen from Switzer- land (Fig. O: 1). With the only exception that the flowers in the American plant are tetramerous, but pentamerous in the European, the structure of the flower and foliage agrees very well with that of the European; and although P. rotata varies somewhat with respect to the structure of the flower: tetramerous or pentamerous, the calyx-lobes being shorter or longer than the corolla, and the leaves varying in length and width, no transitional forms have. been observed so as to make the identification uncertain. P. rotata Griseb. is generally a taller plant with the leaves relatively much longer and narrower than those of the other species, and the narrow calyx-lobes constitute also a good character. As stated above, the flower varies from tetra-to pentamerous, the former observed in the specimens from Bathurst inlet; in Greenland both types of flowers occur, and in a large series of specimens from Amur most of the flowers were pentamerous; pentamerous flowers are also characteristic of the plant from Colorado. With regard to the plant from the Hudson bay region, some small individuals from Fort George have only tetramerous flowers while others of normal size have all the flowers. pentamerous; in very tall specimens from Alberta the flowers are constantly pentamerous. The calyx-lobes vary con- siderably in length as may be seen from the figures (Fig. O: 6, 8, 9, 11, 12 and 13); for instance in some specimens, collected growing together (Figs. 12-13), the lobes may be longer than the corolla, or considerably shorter than this. But in these deviating forms no approach to the floral structure of P. carinthiaca can be detected. : These two species are strictly annual and of low stature; the tallest speci- mens of P. rotata measures 10 cm. in height, while P. carinthiaca does not reach more than from 2 to 3 cm. POLEMONIACEAE. Phlox Richardsonii Hook. Densely caespitose, forming large cushions of a diameter up to 18 cm. The primary root persists; it is deep, woody, and branches very little; the flower- bearing shoots are quite long and woody, bearing numerous crowded rosettes of small leaves at the apex, surrounding the single, almost sessile, relatively large flowers. The plant thus represents an undershrub. Polemonium caeruleum L. var. villosum (Rud.) Brand.! Measured from the crown of the very long and slender persisting primary root, the prostrate stem above ground reaches a length of up to 25 cm., and stems of that length are in their fourth year of age. The plant is stoloniferous, and the development of the shoot may be studied from smaller, younger speci- mens. During the first season a rosette of leaves develops; in the second year a terminal inflorescence appears, borne upon a long (about 10 cm.), erect peduncle, naked, or bearing a single leaf supporting a secondary few-flowered inflorescence. From the axil of one of the rosette-leaves a long stolon develops bearing numerous seale-like leaves and terminating in a rosette of green leaves from which the ramification continues as described. The vegetative reproduction is secured, moreover, by long, slender stolons proceeding from the axils of the scale-like leaves; thus the plant is able to spread over an area of quite considerable size. 1 Polemoniaceae, in A. Engler, Das Pflanzenreich. Berlin, 1907. 56 B Canadian Arctic Expedition, 1913-18 Figure P. Polemonium caeruleum L. var. villosum (Rud.) Brand, showing the stolons, and the base of an aerial shoot; two-thirds of the natural size; specimen from Herschel island. P. boreale Adams. In this species the height of the flower-bearing stem is only about. 6 cm., but the total length of the shoot developed from the crown of the root may measure a length of up to 20 cm. Otherwise the structure of the root and the ramification of the shoot agree with those of the preceding species. BORAGINACEAE. Mertensia maritima (L.) S. F. Gray. The primary root is fleshy, quite thick, and reaches a considerable length —over 30 cm. From the crown of the root arises a large complex of shoots which ramify freely, forming open rosettes and ascending, very leafy, flower- bearing stems. In the specimens from Martin point, Alaska, the plant forms large open cushions measuring about 20 cm. in diameter. , Myosotis silvatica Hoffm. It is rather surprising to see an herb like this reaching a height of 25 cm. at a latitude of 69° 35’ N.; the specimens were collected with flowers and young fruits in the month of August. SCROPHULARIACEAE. Castilleja pallida (L.) Kunth. The tallest specimens are from the south coast of Coronation gulf; they measure about 22 cm. in height. The plant is perennial but the primary root, although quite thick, is not very deep owing to the profuse development of long, thick, lateral roots. There is no rosette of leaves, but a cluster of shoots arises from the ‘ Arctic Plants: Morphology and Synonymy 57 B crown of the root, and most of these shoots are flower-bearing, terminated by the dense, capitate inflorescence; usually some few purely vegetative shoots develop from the crown of the root, but these do not grow any further, nor do they winter over. The vegetative reproduction is effected by means of the ia nage stem-bases with their auxillary buds, thus representing a pseudo- rhizome. Pedicularis [Tourn.] L. With respect to the habit of the species collected, four types are represented. They are all perennial, and the primary root persists, but is developed to quite a different extent. P. arctica, P. lanata, and P. hirsuta represent one type in which the primary root is very thick, wrinkled, and commences to branch at a depth of about 2 cm.; it then develops a few lateral roots, equaling the primary in length, and of almost the same thickness. There is no rhizome, and a compact rosette of green leaves is borne at the crown of the primary root. In the centre of the rosette the flower-bearing stem develops. In P. sudetica, on the other hand (Fig. Q), there is a subterranean, horizontally creeping stem which is FIGureE Q. Pedicularis sudetica Willd., showing the horizontally creeping rhizome, a stolon, some leaves, and the base of the flower-bearing stem; natural size; specimen from Camden bay. 58 B Canadian Arctic Expedition, 1913-18 quite thick. Secondary roots develop here and there; they are fleshy but rather slender, long, and branch very little. Besides that the apex of the subterranean stem develops into a floral shoot, as shown in the figure, some lateral shoots develop also, and these remain vegetative for at least two seasons. By this structure of the stem and root-system P. sudetica is readily distinguished from the three species mentioned above. A third type is shown by P. verticillata, a very conspicuous plant with the several flower-bearing stems measuring about 18 cm. in height. In this species the habit, i.e. structure of root-system and shoots, agrees with that of Castilleja, described above. Very characteristic is the arrangement of the stem-leaves, forming several whorls and subtending~ axillary, almost sessile clusters of flowers. In P. lapponica and P. capitata the primary root is long, but remains very slender, and persists only for a few years. Long, creeping stolons develop from the axils of the basal stem-leaves, and when separated from the mother plant these stolons give rise to new individuals. When I collected P. lapponica in Greenland, I observed that the flowers are very fragrant, exhaling a perfume similar to that of the Lily of the Valley; otherwise the species of Pedicularis are not at all fragrant. In some of the European species the ramification of the shoot has been described as being monopodial; owing to the present material being dried, and all being specimens in full bloom, I was unable to study this particular point in the stem-structure. SELAGINACEAE. Lagotis glauca Gaertn. This has a long (about 10 cm.), creeping, fleshy rhizome, bearing many long, slender, sparingly ramified, secondary roots. A flower-bearing stem develops at the apex of the rhizome, surrounded by a few (mostly only two) green leaves, but by many withered leaf-sheaths from the preceding years. The plant shows to some extent the same habit as Chionophila and, judging from the well preserved, dried specimens, the floral stem appears to be lateral; in other words, the shoot represents a monopodium as in Chionophila. According to Hooker (FI. bor. Am.) the variety Stelleri is the plant collected on the expedition, but Hooker considered this a species, not a variety; and he calls the genus Gymnandra. The specific diagnosis reads as follows: “G. Gmelint (Cham. et Schl.); foliis radicalibus subrotundo-aut elongato- ovatis basi parum attenuatis obtusiusculis grosse inaequaliter crenatis, stamini- bus labio superiore duplo brevioribus, stylo illo breviore. G. ovata Willd. G. reniformis Willd. G. borealis var. Pallas. Lagotis glauca Gaertn. Bartsia gymnandra Willd. Hab. Unalaschka Cham.” “G, Stelleri (Cham. et Schl.); foliis radicalibus oblongis utrinque infra vero magis attenuatis acutis inaequaliter obtuse serratis, staminibus fere longitudine labii superioris, stylo illo longiore. G. minor, G. dentata, G. gracilis Willd. G. borealis Pall. Hab. Bay of St. Lawrence. Chamisso.”’ PLANTAGINACEAE. Plantago lanceolata L. var. Characteristic of this arctic variety is the small size of the leaves and floral scapes; the latter measure generally only 5 cm. But it agrees with the typical plant so far as concerns the structure of the shoot being monopodial, and by the development of the primary root as a deep, thick tap-root, measuring in some specimens about 10 cm., and bearing several, but thin lateral roots. Arctic Plants: Morphology and Synonymy 59 B VALERIANACEAE, Valeriana capitata Pall. This has a long, about 7 cm., creeping rhizome with many slender, un- branched, secondary roots, while there is no trace of the primary root, at least not in fully matured individuals. There is no rosette of leaves, only a pair of basal ones, and the aerial stem seldom bears more than two pairs. CAMPANULACEAE. Campanula uniflora L. (Fig. R). Some remarkably tall specimens measuring 22 cm. were collected on the south coast of Coronation gulf; all the other specimens are of normal size, and much smaller. The vegetative structure is quite interesting. As'shown in the figure, the primary root is developed as a relatively large, wrinkled tap-root, Figure R. Flowering specimen of Campanula uniflora L., showing the large primary root, the ascending subterranean stems terminated by. rosettes of leaves and flower-bearing steams; natural size; * specimen from Bernard harbour. 60 B Canadian Arctic Expedition, 1918-18 bearing several shoots at the crown. These shoots are slender, and their sub- terranean part bears small, scale-like leaves, while toward apex small rosettes of green leaves appear, surrounding the flowering stems. As the main, i.e. the primary, root is wrinkled, indicating that it is contractile, the crown of the root becomes gradually pulled deeper and deeper down in the ground, at the same time as the subterranean parts of the stems increase in length. In young speci- mens the subterranean stems are thus much shorter, sometimes barely visible, the crown of the root being close to the surface of the ground. COMPOSITAE. Erigeron grandiflorus Hook. A robust species, the heads measuring 4 cm. in width, borne on simple, leafy stems, about 12 cm. in height. A large rosette of leaves is developed from the short, thick, ascending, subterranean stem portion, and the primary root is deep, but rather slender; similar long secondary roots develop from the subterranean internodes. In large specimens several subterranean stems, terminated by leafy rosettes, were developed from the crown of the same root. A corresponding structure recurs in EH. compositus Pursh, but in this species the subterranean stems are more numerous; thus the rosettes form cushions of considerable width. E. uniflorus L. In well developed specimens the habit may vary from densely caespitose, i.e. with a rosette of leaves borne upon a very short, erect, subterranean stem, bearing numerous strong, secondary roots, to a more open growth, the sub- terranean stem being horizontally creeping, measuring about 7 cm. in length, and terminated by a few aerial shoots with only a few basal leaves. Most of the species belong to the variety pulchellus Fr. E. alpinus L. Only a single, but well developed specimen of this, on this continent so very rare species, was found at Port Epworth. As pointed out by Lange (Consp. FI. Groenl.), it may be readily distinguished from E. uniflorus L. by ‘‘Periclinii | pilis non vel obsolete articulatis, floribus femineis exterioribus ligulatis, interiori- bus oblique tubulosis.”’ Concerning the habit of the plant, this differs also to some extent from that of the preceding species. The basal leaves form a compact rosette, with numerous strong secondary roots; from this rosette nine erect flowering stems were devel- oped, all simple, and attaining a height of about 16 cm. Antennaria alpina (L.) R. Br. Only a few specimens were collected and they represent the variety in which the leaves are hairy on both faces. Lange? has called this variety canes- cens, and it has been found in Greenland, Scandinavia, and in the boreal regions of this continent, besides in the alpine region of the Rocky mountains, in Colorado for instance. A. candida Greene. The specimens upon which Greene established this species came from Mt. Le Washington, at an elevation of 9,000 feet, and the diagnosis reads as ollows: 1 Flora Danica Fasc. 47. Tab. 2786 (1869). 2 Greene, E. L. Leaflets, Vol. 2. Washington, 1910-12, p. 151.. Arctic Plants. Morphology and Synonymy 61 B , “The numerous very. leafy surculi upright and their leaves not rosulate, but suberect, 3 inch long more or less, obovate-spatulate, clothed densely with close snowy-white tomentum: flowering stems rising 1 to 2 inches above the cushion of leafy surculi, slender with their small leaves as white-whooly as other parts; heads in fertile plant 3 to 5, closely congested; proper scales of involucre wholly concealed by the white indument, their translucent tips from oval and obtuse in the outermost to oblong, and in the innermost to lanceolate and acute or acuminate; sterile plant not known.” __ The material from Bernard harbour shows an ascending, creeping rhizome with numerous long, slender, unbranched roots; the foliage forms dense rosettes but so closely crowded that the leaves, as described, are mostly erect. The snow-white tomentum on both faces of the leaves renders the species very distinct from A. alpina and its allies. Matricaria inodora L. var. grandiflora (Hook.) Ostf. This is the variety which Ruprecht! has described under the name “‘phaeocephala.”’ It is the arctic representative of M. inodora, and while the typical, southern plant is annual or biennial, the arctic form is perennial. The specimen from Herschel island measures about 23 cm. in height, and only a few specimens bore two heads on the same stem. The subterranean stem-portion is erect, but very short, densely covered with remnants of old leaves, and the root system consists of long, slender, fibrous roots. Several rosettes of green leaves may occur at the base of the aerial stem, and the habit of the plant is actually that of a perennial. ’ Chrysanthemum integrifolium Rich. Characteristic of this species are the short, linear leaves forming very com- pact rosettes; the flower bearing stems are monocephalous. The subterranean, vegetative organs consist of several ascending stolons arising from a common base, evidently the crown of the primary root which, however, had faded away in the material collected; the only roots still in existence are long, slender, second- ary ones, proceeding from the internodes of the stolons. Artemisia vulgaris L. var. Tilesii Ledeb. Some very tall specimens were collected on the south coast of Coronation gulf, which measured a height of 40 cm., including the about 12 cm. long panicle. The aerial shoots are developed from a complex of stem-bases with numerous, thin, secondary roots, representing a pseudo-rhizome. A. Richardsoniana Bess. and A. hyperborea Rydb. These have a persisting, deep and thick primary root, from the crown of which ascending subterranean stems arise, terminated by compact rosettes of leaves, surrounding the flower-bearing stems. These subterranean stems vary much in length, averaging about 6 cm. in the latter species. .According to the diagnosis the hairy covering of the stems and leaves is somewhat different in these species, which, however, may depend on the character of the soil where they were collected, the former inhabiting a gravel-tundra, the latter, on the -other hand, sand-dunes and sandy slopes. With respect to A. Richardsoniana Bess., this is described by Gray? as follows: ‘A span to near a foot high, with rather slender ascending stems from a cespitose 1 Samojed. cisural, l.c., p. 42. 2 Synoptical Flora of North America. The Gamopetalae. Second Edit. New York, 1886, p. 371. 62 B Canadian Arctic Expedition, 1918-18 caudex; leaves silvery-canescent with firm very close-pressed pubescence; radical twice ternately or quinately divided or parted into oblong-linear or narrower lobes (of only 2 or 3 lines in length); cauline sparse, mostly trifid; heads comparatively small (2 lines high), several or rather numerous in a strict and simple racemiform inflorescence, fuscous; corolla pilose or sometimes glaborus.—A. arctica and A. caespitosa, Bess in Hook. Fl. I, 328, 324. Arctic coast to Bear Lake.” A. comata Rydb. In specimens from “Sandspit at Martin point,’’ the numerous, dense rosettes of leaves are developed at the apex of short, very much branched, sub- terranean stems with an abundance of thin, profusely branched secondary roots. In specimens from “Gravel-tundra at Collinson point” there are long, subterranean stolons, quite thick, but with only a few roots, which are quite slender; in such specimens the cushion-structure is less pronounced. Petasites frigida (L.) Fr. This is one of the species which by several authors have been referred to Nardosmia Cass. on account of the corollas of the pistillate flowers being ligulate. By Reichenbach Nardosmia has been reduced to a mere section of Petasites, and so it is accepted also by Bentham and Hooker. Among the species of the section Nardosmia, N. glacialis Ledeb. and N. Gmelini DC..have the basal leaves developed almost contemporarily with the flowers, while in the other species as well as in Petasites s.s. the inflorescence appears earlier than the leaves. ‘From a morphological viewpoint the genus Petasites offers several points of interest, for instance with regard to the structure of the rhizome; in this respect several species have been discussed by Hjalmar Nilsson, viz.: P. alba (L.) Gaertn., P. spuria (Retz.) Reich., and P. officinalis Moench. In these the rhizomes bear series of membranaceous, sheathing leaves, destitute of blades, and some with blades and long petioles. In P. frigida the following structures have been observed. In a fruiting specimen from Herschel island, collected in August, the horizontally creeping rhizome bears a tall (22 cm.) flower-bearing stem, and at the base of this the rhizome continues as a horizontally creeping stolon, consisting of six internodes. The first three leaves are scale-like, but after these follow two long-petioled, green leaves, situated very close together, while the leaves of the apical portion of the rhizome are merely scale-like. A fragment of a stolon from the same locality (Fig. S: 2) illustrates this structure. In another specimen (Fig. 1) from the south coast of Coronation gulf, collected in July, the rhizome bears a tall, flower-bearing stem at the base of which two young, green leaves are situated. The apical portion of the rhizome bears four scale-like leaves, of which the foremost partly surrounds two, very young, green leaves. In other words, the rhizome is undoubtedly a monopodium with the flower-bearing stem axillary, as shown in figure 1. Moreover, it would appear as if the green leaves (Fig. 2) represent a very short vegetative branch, developed in the axil of a scale-like leaf. Concerning the root-system, secondary, long and slender roots proceed from the internodes; they are especially abundant close to the floral shoot. The fact that P. frigida very seldom develops flowers in the arctic region, and that it must depend, to a very considerable extent, on the vegetative repro- duction, seems to indicate that the species is not well adapted to the arctic 1 Dikotyla jordstammar. (Acta Un. Lund. XIX, 1882-83, p. 179.) Arctic Plants: Morphology and Synonymy 63 B climate. Kjellman ! has expressed the view that the plant like several others reached the arctic countries at a time, when the climate was more suitable to its existence. Nathorst has reached the same conclusion relative to its occurrence in Spitzbergen. oo ' Figure Ss. 1. Petasites frigida (L.) Fr., showing the rhizome, the leafy shoots, and the base of a flower bearing stem; natural size; specimen from Tree river. 2. Same species, showing the matured foliage; two-thirds of the natural size; specimen from Herschel island. 3, 4 and 5. Basal leaves of Senecio resedifolius Less.; natural size; specimens from Herscte island. Arnica alpina (L.) Olin. Several very tall specimens, measuring from 25 to 30 cm. in height, and bearing three heads were collected at Bernard harbour and on the south coast of Coronation gulf. There is a horizontally creeping rhizome of considerable length, densely covered with remnants of leaves, and with many slender, un- branched, secondary roots proceeding from the lower face of the internodes. At the apex of the rhizome a few pairs of opposite, green leaves are situated sur- rounding the base of the flower-bearing stem. Characteristic of Arnica is that the rhizome bears only green leaves, none scale-like. 1 Ur Polarvaxternas lif. (l.c. p. 504). 64 B Canadian Arctic Expedition, 1913-18 Saussurea alpina (L.) DC. Has a horizontal, woody rhizome with scale-like leaves, but the root- system is very scantily represented. A dense rosette of leaves terminates the rhizome and surrounds the floral shoot. Crepis nana Richards. This interesting little plant forms dense cushions from only 2 to almost 10 em. in diameter. The primary root persists as a relatively thin, but deep, persisting root, branching only near the very slender apex. In continuation with the root an erect, subterranean stem arises, bearing numerous, long- petioled, green leaves and numerous, very short flowering stems which have. from two-to four heads. The subterranean stem is quite thick, and shows very plainly scars from withered leaves, crowded on account of the internodes being extremely short. Arctic Plants: Geographical Distribution 65 B t CHAPTER II. GEOGRAPHICAL DISTRIBUTION. By examining the geographical table (Table 1, pp. 68-75), in which I have enumerated most of the species which were collected on the arctic shore of this continent, we notice that this flora is actually a composition of two, which are very distinct, viz.: a northern and a southern. “Of these the northern will, by a close analysis, soon prove to contain a large contingent from districts very remote and, although mainly northern, by no means to be restricted to these regions. But we call the flora “northern” because the most conspicuous part of its components represents the circumpolar flora. With regard to the southern element, this we readily recognize as being principally derived from our own flora, the components being either identical with American types, or at least exhibiting a striking analogy with our native species. Such dual composition recurs in other arctic countries, marked to a greater or lesser extent, very characteristic in Greenland, ‘less so in Finmark and Lap- land, Spitzbergen, etc., but quite distinct on the Siberian coast, judging from the interesting and highly instructive sketches contributed by von Baer and Kjellman. For even if the arctic Siberian flora is not so very distinct from the circumpolar or the truly arctic flora, there are certain types which may readily be recognized as being of southern origin. Of great interest is the fact that even on the ‘‘Barren Grounds’ are species which also are at home in more hospitable surroundings, the Alps of Switzer- land, the Pyrenees, and the Himalayas. And the enormous distribution, ex- hibited by some of these plants, may solve the problem as to the migration of the ancestral, circumpolar species during the glacial epoch. But at the same time such widely scattered distribution farther south may lead to the acceptance of the theory pronounced by Schouw who, with so very few words, but substanti- ated by brilliant observations, explained the problem as “‘Eadem ‘momenta cosmica easdem plantas diversis in locis produxisse.”! In other words, while emigration and return of the arctic plants, so admirably discussed by Nathorst,? cannot be contested, there is nevertheless some difficulty in realizing the fact that some of these northern types have found their way to the mountains at such enormous distance from the arctic region. Nevertheless, some of these difficulties have been removed by the many important points which Nathorst has brought up, especially with reference to the supposed composition and distribution of the former arctic flora, its migra- tion to the south during the glacial epoch and, finally, its retreat when the ice~ receded. It is, for instance, an indisputable fact that the Altai mountains harbour a considerable element of the arctic flora besides which the vegetation of these mountains shows an unusually large number of species common to other mountains, though far remote. And the natural cause of this wide distribution of the Altai flora is by Nathorst explained as depending on the favourable con- ditions in these mountains during the glacial epoch. When the temperature decreased, and the alpine element of the Altai flora was forced to descend to the lowlands, the flora was then able to distribute itself over the lowlands of all northern Asia. For, contrary to Europe, there was no inland ice to prevent its distribution; the plants were enabled to spread as far north as the climatologic conditions permitted. Moreover, the relatively northern situation of these mountains was an important factor with regard to the migration of the plants to the arctic region. ; . 1 Schouw, J. F. Dissertatio de sedibus plantarum originariis. Havniae, 1816. ; ; 2 Polarforskningens Bidrag till Forntidens Vaxtgeografi. (4. E. Nordenskidld: Studier och Forsknin- gar. Stockholm, 1883, p. 231.) 24657—5 66 B Canadian Arctic Expedition, 1913-18 No doubt the Altai mountains constitute a most important centre of dis- tribution, and many species were developed there which succeeded in reaching the northeastern corner of Asia, and from there extending to arctic America and Greenland. A similar important centre was undoubtedly located in the European Alps, known to be so very rich in alpine species, many being endemic to these mountains. Furthermore, the Rocky mountains bear evidence of having been the actual centre of an old, alpine vegetation, of which several species have entered the arctic region and now constitute a large element, a truly American one, of the polar flora. ; These data are consequently of great importance to the solving of the problem regarding the probable location of the centres of development and distribution of our arctic plants. : However, such singular distribution of plants is more readily observed, when we deal with the alpine vegetation, comparing this with the arctic. For instance, in Colorado the alpine vegetation illustrates the fact at once, that it is composed of six elements: circumpolar types; arctic, but not circumpolar; northern, but not arctic types; northern types, endemic to North America; and finally there are some southern types common to both Worlds, or endemic to North America. This heterogeneous composition naturally induces us to attempt some sort of tabulation of the complete geographical distribution of the arctic species. But it is by no means an easy task to mark down a number of species credited to the various districts in the north and south. And the difficulty presents itself from the indisputable fact that many species reported from the arctic have also been recorded from the south, but under other names. I think especially of such species as have originated in the arctic region, but were forced to migrate to the south during the glacial epoch; some of these were left on the southern mountain summits, while the others returned to their northern homes when the ice receded. Such species as were left on these moun- tains thus represent remnants of a glacial flora, and naturally the change in the climatologic conditions has, sometimes, resulted in their appearance under disguised forms which not infrequently pass for different species. Nevertheless, if such tabulation be not absolutely correct, it will always give us some idea of the distribution, in a general way, of a number of species which are still recognized as identical, northern as well as southern types. Having thus reached to obtain a view of the geographical distribution, the alpine element proves a surprisingly large representation in the north, and especi- ally in the arctic region. And besides that, this same vegetation, at least a number of the alpine species, show also a more or less extensive distribution through- out the southern mountainous regions of both Worlds. Among these species, we meet thus with some that extend to the arctic region, and others of which the distribution may extend far to the north, but without reaching the arctic countries. Of these the latter frequently show a vast distribution throughout the northern hemisphere and, moreover, a distribution which may prove . exceedingly scattered. The occurrence may be so scattered, indeed, that we are at a loss to appreciate how the species ever reached such stations, so remote from each other. And as will be demonstrated in the subsequent pages, the species, which I have in view, are not only such as occur on the summits of the higher mountains, very remote from each other, but also such as are more or less lowland plants. With respect to the origin of some of these, I have suggested the probability of more than a single centre, which by no means will be in opposition to the indis- putable theory relative to the migration of the plants during and after the glacial period. Some few examples may be cited to illustrate the probability of a single centre, viz.: Carex festiva, and of evidently several: C. atrata and C. alpina. Having studied several species of the genus Carex from this particular point of view, we might consider for instance C. festiva. This species is arctic, but neither Arctic Plants: Geographical Distribution 67 B circumpolar nor strictly alpine; it is relatively rare in the polar regions, but it has been recorded from northeastern America, Greenland, Iceland, and Finmark, where it occurs only as what may be termed the “typical” plant. But much farther south and especially in the subalpine zone of the Rocky mountains is a herd of this same species, accompanied by several aberrant forms, besides by species that are apparently distinct, but among its closest allies: C. athrostachya, C. pratensis, C. petasata, etc. Judging from our present knowledge of the distri- bution of C. festiva, its geographical centre seems to have been in the south, in the Rocky mountains, where it is, thus, typically developed, and associated with characteristic varieties, and with allied species. A more complicated case ' may be illustrated by Carex atrata and C. alpina. The former is only known as arctic in a few stations of Greenland and the European continent, while the latter has been collected in arctic Russia, Finmark, Greenland, and North America, but is more frequent farther south. In the Rocky mountains of Color- ado these species are associated with several allied types, e.g. C. melanocephala, C. bella, and C. chalciolepis; a close ally of C. atrata, C. ovata, abounds in the northeastern part of this continent, thus illustrating the occurrence of allied types associated with each other. This might indicate the location of the geographical centre as being in the Rocky mountains, so far as concerns the American representatives of C. atrata and C. alpina. In Europe C.-atrata is, to some extent, accompanied by two plants, C. nigra and C. aterrima, both of which may be looked upon as immediate allies of this species. And if we extend our comparison of these species with those that occur in the Himalayas, we find there not only C. atrata and C. alpina, but also some deviating forms, and some allied species, among which C. Lehmannii, C. obscura, C. Duthiet, and C. nivalis. If thus the association with allies in connection with frequent occurrence and tendency to vary may throw some light upon the question as to their centre of distribution, I certainly consider it natural to attribute to these species three distinct. centres, one in the Rocky mountains, another in the European Alps, ‘and a third one in the Himalayas. Similar results may be obtained by comparing the distribution of other plants, when represented in both Worlds. For instance, Anemone narcissiflora, Lloydia, Elyna, Pyrola, Saxifraga, and several others from the arctic shore are very instructive from this particular point of view. 24657—54 68 B Canadian Arctic Expedition, 1913-18 Polar Regions 3 g u nQ pad 2 a |e om on) mn ag 5 + faa} a io a n a 8 - = 25 5 $ a ete | 2 ae a be Ss i ‘3 le 9 Geographical 23 gi) a/4/4 i es ales F wsi_falglalela] lel [@1./4] 2 (Bel 2 Table 1 SSlay]}S/ 81s] 8] 2 A 8 s 2|e'a| 2 cl al e(e le) S| a 3/8|5.| 2) 8] # lgsl 3 nO) ae} aq Rilo! = 3S > |. 2a] 3 Sel elelsisle/8] a2] sis] ea] 2 Sal 3 aa] o/ 4 | oO] 5 rPlolal dal 31.88 E/s |e s| 2 ' a EIB etTa] ©] a] el] Sal are" eis) 3s B-Al eo 2 lolalaidi/2zl4/Sj/4jolea |B)/a);e ja |< Cystopteris fragilis (L.) Bernh.) *| *| *| *] * lec ies (a a Dryopteris fragrans (L.) Schott} * ae Equisetum arvense L........... <0) md ae (A oe ee E. variegatum Schleich........ sid Pt | a Dd =| # mg ee * Lycopodium Selago L.......... il Late c(i ca a | * «| # Hierochloé pauciflora R. Br...| * mt oe » iad eee + *y: lama? H. alpina (Lilj.) R. et 8....... * * * * * *] * * * * * Alopecurus alpinus Sm........ * * * * * * * * Arctagrostis latifolia (R. Br.) GTS ee ostatagrasnarasecee aera si) ta We 2 * * * Calamagrostis purpurascens R. IB Es sce denne teen atcoweiNe ph l|| = seal Trisetum spicatum (L.) Richt...) *| *| *} *} *] *] #7) *] *] *P FP FL Fy * «| * Catabrosa algida (Soland.) Fr..} *| *| *| *} *] *] *] * * «| o# Poa glauca M. Vahl........... | OR SRR RE OR a Mes acslhtatesied | * * | * P. abbreviata R. Br............ bid Nee Ole P. arctica R. Br............5-- BOT BRL a ea AR x] oe] oe] oe] * Arctophila effusa Lge.......... Th 4a Sins My Fh * * Dupontia Fischeri R. Br......- ¢ * | ep e * Glyceria vilfoidea (Ands.) Fr...) *| *| *] * |...) * |] * 1]... G. tenella Lge.......--...0-00- Me lx ai le ssccceea ave lekevensal| ° G. vaginata Lige.........-.6056+ Pe ss ~ | * Festuca ovina L. var. brevifolia.| *| *| *| *) *] *] *] *] FL *) FY FTF «| «| x F., rubra L. var. arenaria....... * ee Oe * * * * + * * * * * * F, altaica Trin................ edie LL Gel noc [teal | cairns asideat hay wade Ilo te| Dataadieoavel * * * Hordeum jubatum J........... sea Deans Pearce een caeel eee DAP asiee mentee At gw facgern boiigye hedged Op) % Elymus mollis Trin............ ~ oe * * * Eriophorum Scheuchzeri HOGG 8 occocnxti areceuracreys is ss * *] * * * * * * z a E, vaginatum L........- 060.06 sal ania a [Perea ean Wiel (es ce a Mol * * E. angustifolium Roth......... * * * * * * * * * * * é a Elyna Bellardii (AMl.) Koch...) *| * J...) *] * [0.0 *] *] *] * “| me |e Carez incurva Lightf........... at ow poe [ow > oe, oe foe! oe] ox] x * Pa ie C. rigida Good.........-. 0000 * * * * * * * * * * * * * * * * Arctic Plants: Geographical Distribution 69 B . Polar Regions 20 as 2] Js| JB 14 : = 4 a 3 | ale i Geographical ae 8 ® ae eS Sa Algal ala o 6 4| 8/8 | ¢8 Table 1. as|_l|elelel|s|3 5) ls $) 8 \8.1 3 Sele Bl Sle ele] (Sl elglele| eleka Bee eleleiS (Solel al slea4a| 2] 2 lfe]3 Ba) 8/28)2)2/8|2|4|'s| 2 (88] 2/3] 2/23) 2 2||8l|4l4j2/4|2/2|62"|e| 9 a aes | 3 Geographical se a e 3 a | -& lo 3} : Sa Sla@|@|.s 2 i alei2 | & Table 1. 3814/6818 4 F 3 c a as ve Sz a Se Bl elelelel_|e| 2/8.) Sle] elgz4 2g| S| 2 N o/s] @lo#| $/ 8] 1288] 8 : Gel se) 2/8) 8] 0/8) 8] 2] $/S8) es] | 216 ta aqj S/2/5/8/5/3/ 4] 8! 8 |Se] 812] 8 le8] & a ~ + _ n rt. — nes 42 /o|/@/4)4]l2)/4/8)/s]/o]j4 |e] 4]ea le 4 Betula glandulosa Michx....... meee. =f * ba eee Polygonum viviparum L..+..... * i a * =| * * * % ie * * bg ba BP BistOnta Un cccnstade dined Suu il eee alleen somes | Pah PO ac (i * 5 Rumez arcticus Trautv........ : * * - ~ Oxyria digyna (L.) Hill....... Se) OE ae ae ee ee oe Re a ee Silene acaulis L............... DS aN oe ae PED a ae 1) 38 id 2 ep Lychnis apetala L............. ale * * * * * * * + al ee BG Hinis RABI sac tapas ches ico | L. Taylorae Robins............]....].. Stellaria longipes Goldie....... ey ee |e | eux * a) & S. humifusa Rottb...........- Ad ee ce Oe ee De * | * Cerastium alpinum L.......... oF) aka ae ae ee ap ae ow +; 1 * ods C. marimum Le... eee Isisoad tate * * * . Halianthus peploides (L.) Fr...) *| *| *] *] * | *] #] * * * Alsine arctica (Stev.) Fenzl...|....| *?). * 3 * fF A. verna Bartl. var. rubella Lge| *| *| *| *{| *] * | #] # * * *] * A. macrocarpa (Pursh) Fenzl...}....|....]....)....]....]....] * * * Merckia physodes Fisch........ Whi |B toa olleck agile abcen’ Visual me * Caltha palustris L. forma radi- cans (Fr.) Hartm........... - *, #) 8) ew] wel oe] oe] wl el @ «| os Aconitum delphinifolium DC ..|....]....)....J....]....[....) * * * Delphinium scopulorum Gr. var. glaucum Gr............ slcitecal| ete aa Tle ae eal nds [ayestnallancva so] ih acs * * Anemone hirsutissima (Pursh) ae MOMAN ciacssansoxcnee-ccisace ive ctaea|'s boreal yaei valid «| o* A. Richardsonii Hook......... a * * * A. parviflora Michx............ RA As ell set cavaalhe decd vave-adl acta | penreasl ieee wal seed. oa A. Drummondii Wats.......... sanded pie * Batrachium confervoides Fr.....|....| * * Wien Ranunculus Pallasii Schl....... cavaleeeel ao cw| e + R. Purshii Richards........... obi % a] ox R. Cymbalaria Pursh.......... aca, ™ ef oat oe] R. hyperboreus Rottb.......... OR Ra a AR RE cal ied RE Balak [tee Arctic Plants: Geographical Distribution 71-3 ' Polar Regions 20 2g 24 r| E s |4 ae a . i a FF Q g a > Q 3 Geographical ae a 3 a a f= lee ie Sl/alala| [@l |e a) 3/9 | 48 i mn fo} sec siei 8] 2) 2lele] [&| 18 | 2/8] 8 leal 4 -a/| oO i n ]o cel SB) = a | a | ao] > q zg gea/2/e|eiS/S/e] 3] glea] 2] 2] 2233 Bal s|s/el/e/e/3) ala] see glal isu e 4<4)2)/2/2/8) 8/8/30] 8) slau) 2/3) 8 igs 2 42 lolalad/dial4]/e8/slole |e] s]a le] 4 R. pygmaeus Wahlenb......... Be RS oe oe * * a] * R. lapponicus L............... Pp A AE eS eed ee | Se wi R. gelidus Kar. et Kir.}....... adasiaall| 4 seat ig cars| cog all Saeed ital ease lees al eae [+ Pole esau glaseeiasaen| tere deat R. sulphureus Soland.......... Se) ok *| * * R. nivalis Li... ecco BR Re ae oe) ae 4 | + * R. affinis R. Br............... Soe iia as x) x a R. Sabinii R. Br............... ete fone R. occidentalis Nutt. v. robus- CMS oes dsneertea o's yscasere diel citesieee eat abst Papaver nudicaule L............ sil is eecatel (saa (cc | | +] o«] «] *] en Lesquerella arctica (Rich.) ‘ian toc ionwan aeons Heme eae ee ™ aes perete | Mme eee Thlaspi alpestre L.v. purpur- i ascens (Rydb.) Ostf......... staslaoyd|as esl edeadens ee iat is Cochlearia groenlandica L....... eee) A || ae ae Draba alpina L..... 6.6. eee lal (coll acid leas (ial oll cl id *) oe] * D, nivalis Liljebl.............. * i 3 * ig ie * * - shee D., fladnizensis Wulf........... ee | ep ae) Rh ae ae * «| oe] # * D. corymbosa R. Br........... geen © | * Deft Mica cieiadasen densa des *[ *) ©) ©) *) #) &] « * oe ee Braya purpurascens (R. Br.) RITA Os cc rating sh ovine le aha av Sye oan |S ol oe Bll 3k eae B. alpina Sternb. et Hoppe....}.-..| * * wi) ® ; * * Eutrema Edwardsii R. Br...... Bl a bis teas |e soe ael| . * Hesperis Pallasii (Pursh) T. Gin shoe oupaae wale ones 3) a atieh Cardamine digitata Richards...) * ¥ seeel C. pratensis Lu. .... cee eee eee Be RO RR AA SEY a) ae a a | RD aR *|] * Parrya macrocarpa R. Br......|--..]--..} * coil isc | *{ o*| # * P. arctica R. Br...... eee ease iain (cee) Paraee cet aren Erysimum inconspicuum (Wats. MoM ilegicsccc esac ouf tues |r e|so ca] site afooos |e ccalinns alossellegiefeagyleeesbaiegtaoas| Sisymbrium sophioides Fisch..|....|....)....)....)e.0.feee fees fees [eee [eee fern fees peee dee feee Pees Chrysosplenium tetrandrum Th. Does /adace dass acetone i ctedotaves tea suas te ese ¥ oe, Fp Sn et Se * x|o* 1 Described from specimens collected in the Ala-Tau Mountains in Central Asia. 72 B Canadian Arctic Expedition, 1913-18 Polar Regions 25 r iE 1S eal] |. | |2/_2 |3 2 $ g| |8 Elels | Geographical a £ s i S ‘é 3 ie Z 3/2/32) 5 3 alels Tables sala /8|/3/2iel8| [@] 12 | e/] sleala e542) 3(2 (= (8/2) ~/ 8] glee] 2] 3/5 es] 2 Zelglelalgle/2la| 2] 2igzlz|a|2les| 3 S<4|2/2/E/E/S12/s|B| 3lez] 8/3) 8 jes] & 4S 1Ojal/dqle¢/a2s/a4}/S}/4)/0l4 | Ala] ajc [< Parnassia palustris L.......... eeitllie eeealea’gial| | a] AR) ATL EY ee, Se Soh Sl am P. Kotzebuei Cham. et Schl...].---| * * ba leas Sazifraga radiata Small........ 2h dccdeell ca eal sodreal cata Heceava| OF * S. rivularis L.......6. 6060-0 cam) (el ie ce (i * ie S. cernua L........66 00. eee eee sa el ec a (ee A ( | | al (| 2a) (Ro S. Hirculus L..... 0.60.0. e eee a - * ? = hai ilies * bt * * = * - . em S. aestivalis Fisch. et Mey rr areal vival co calens vlege tual © * ee Pan SS. Nelsoniana Don............ sculls ted ae sca lhegee cae oP NG cooley a etn] ealallece cere lianwe calle Taagiail, eck S. refleca Hook.............. ae SB. nivalis Wis ccndiorscsernancee ce ee ee *| #1 % S. hieraciifolia W. & K........ oa Rn es Oe S. decipiens Ebrh. v. groen- a LENCO eoiecterees eink sido hanes | al el |W * 5g S.bronchialis L................ * * * el ell BE | S. tricuspidata Rottb.......... Fey MEA ese cll kang-ccoal eae | ay avelp ace seyail ovses fer feuthans | wBegeee [leeas| a Saalhewe aca or oe he BIRO ES Lise s yaen s vewn onal Re | eal Elev: AR BE Vent etl coe ea ft re * S. flagellaris Willd............ ee Bleed iate al Fl OF eee ell & * * * * * S: opposttifolta Lie. cci ees xs xa, a a cn a OO Sedum Rhodiola DC........... re a 2 Sie ee Dryas octopetaja L............. ile (iia ac (eed Palacae (iia (ace Pace (2a) cl VO Lacs Dy antegrifolta Val vcscesrncaccel Fi] % [ice calleeg-asreea|sace|esesleecstogaslegeslt ™ locccliacct ae Sieversia glacialis R. Br........ aay e ane ro wllae A daealeeeed |! 8 keeles dled ee ee ae Eee Potentilla palustris (L.) Scop...}....J *]....) *] *] *] *] *] #]cd o* tod *] * aie Py frutioosa Lisciiieacnceiied ones ia Re | covllcss dlat-ae emcall, OF laevolaxace| * * * a ea = PAN, Bticcstacsseaceel | lal vscclincacllecial senglomeition lencl. 1 alec. Pi Ve Masset escovnansedaieeecaestea || * * ie * *i| | * os ee * * * P. rubricaulis Lehm........... * eee ae oo P. Vahliana Lehm............. *| * ede eee P. emarginata Pursh........... *| * «| * ' ae ae Rubus Chamaemorus L........ «| «| *] ok] x * * RL shoves * Rosa acicularis Lindl.......... ; dele eae © ea icloswdlitgaliovenlaeodlatetlancn|) S Lupinus arcticus Wats ers Arctic Plants: Geographical Distribution 73 B Polar Regions Geographical Table 1. American archipelago strait American coast of Ber- Islands of the arctic Greenland Spitzbergen Arctic Scandinavia Arctic Russia Nova Zembla Arctic Siberia Alps and Pyrenees Caucasus Himalaya Altai and Baikal mts. Rocky mountains ~ ing strait Iceland | Asiatic ccast of Bering | Appalachian mountains L. nootkatensis Domn.......... Satie [acans. | aswwaitllhe sabes] wcatadelleneeien a] eieeelle degs| ara aaallag se a|ietedtaalleaenlaeedl o* Lathyrus maritimus (L.) Bigel.}..-.) * |...) * | * fo.0b.} * fp bpp Astragalus alpinus L........... Ledss|aeec| PA oR See) Ry EP EY] oe eae A. aboriginorum Richards..... sts illo eats | esas tel lao anal eva. &| Saces cll span | eee well eae olla See malfesesel | ~Phaca frigida L.........0e ee wefeeeedeee fee FPP HT EY RL Te] Ld * Ozytropis campestris DC...... Wlostal easiest | BE ee eb pee salen nt est | O. foliolosa Hook.............5 ier Ad rhe AM sacar [ts ecnatlee espe aye Al uae alllscoess' llarghiendl won ol Set. ang[ smseeallleostau'le wavaleaees O. Roaldi Ostf..............086 fe ill sda Pes cedle gen Pano laha dl nehliGeesbeewelecadeaaely O. nigrescens (Pall.) Fisch.....| * |....]....[....]....)...., * }occ.d....[..eef # doe..[eceebee eed * O. arctobia Bunge.............. Hedysarum Mackenzii Rich....| * |....)....)....p..0f.00-; bpp pee bP pf # H. alpinum L....... 60. eee eee Empetrum nigrum L..........- *]oe foe) ef oe} ef ef ef et we] ef] | we] e] «| « Epilobium latifolium L........ ss Pe ee E. angustifolium L............ loaf tow] el ot el oe] mw] ae] ew] we] oe] a] « Hippuris vulgaris L........... eae te EE Poel OR] RB Rp cae ewe ae] ae oe Bupleurum americanum C.&R.).-..)....[....J....[--.-f--.- fee [eee peepee bed peep a1 nee Selinum cnidiifolium Turcz..... wees ead ti ae eed G2 awed een Poagaledavibten olvey Aboeet ot Pyrola grandiflora Rad........ Po licen igareleesal Satna cheageleteelen sal lkipe viene) ] Ledum palustre L.............. | Be Pla evel Sal Mee tlle eal Poe lhe ae Rhododendron lapponicum (L.)} Wallis ars nsneny tenia chess Be OE Meets WL Ml ela ltnaie ofioeneral tage opatemetleeeeltay all Sl OR) ae Kalmia polifolia Wang......... Pare ted (ae verr] [ere gel eaeeere||nverees| Pron] | rari Cetra Weesarem [Brann eon | ereen, USN URE * be Cassiope tetragona (L.) Don....| *| *| *]| *] *]....J * Jo. pep * hope? oF] * Arctostaphylos alpina (L.) Vaccinium caespitosum Michx..]....|....}....}....J-...Je00-feee V. uliginosum L.f. microphylla Lge....... jiatiesadetes Haun 3 Per ae 2 V. Vitis-Idaea L. var. pumilum Dodecatheon frigidum Cham. et BGi lier se peniesaksawseceans eed A aeale a abe cod wemalhewes ea bee ee poocfeefoe Canadian Arctic Expedition, 1913-18 74 8B ra Polar Regions 80 28 . ae og 2 oO =] “ ao cs 2 : Q 8 ay = n o Del = ow > 2 faa) "4 4|° ie ' ae a S 3 “4/8/23 ]¢s Geographical caar.| =] a | S| 8 q ‘ales gi als} a|)2 15 Pp % alals |4 Table 1 Saiz} ele} 3] 8/2 a 8 |slulele |e » 4 Ss a S| a © | ca Es} a = > Ble 26/4 | 3 | @ N/@/g] 2] gloel 4] 3! . |eal-8 welaloi/2i/e)/ale] a] 2] sls'sl || 2 les] 4 eElSis|/S/S/E/8/4] 2! 8 (esl 2] 3] 3 les 2 Se) 2/E/S/e1 5/8) 8] a] & laa) 2] 2] 8 eal es 42 lolaldai¢a/2i4/8i/alol¢eq |Hlaleale ja Primula borealis Duby........ ‘s ss P. stricta Hornem............. Seka! Ae sl et a ? = P. sibirica Jacq..........2.0065 anaellasie dbase || Oe * ore a Douglasia arctica Hook........ bd ail bss Zl tan lacansaast he ine S65 Androsace Chamaejasme Host..| * ih iat dl a (i cal ea (ad A. septentrionalis L............ Mh ostomy. So) Me | eT ee ed ST Statice Armeria L. {. sibirica (Turez.) Simm............ ae * ic Wena * ~ * Gentiana arctophila Griseb.....|.... G. propinqua Richards........ Pssieclaw apriie Mlccdeelsuimiatimeses| ® * Pleurogyne rotata Griseb.......]....] * * s * ng ._P. carinthiaca Griseb.......... ns fisoy pall Bt 4 “sssasvo' eo aeded i “) ® Phlox Richardsonii Hook....... * Ree Polemonium caeruleum L. v. villesam (Rud.) Brapdecve:lexs : g ~Q =| aie 2 : Sg a FA 3 glale |g Geographical car] si/a/| | 5) 2 3 3 ae 3 g|e|¢/2] 8 2 Q a} eie | 8 Table 1. sit} S| s)/5] 8/2 aA) Is | sl alg | ak) a | 3B|/2l/H/Q]a 3 Oe eas B a $42] 8 gel elie] eles eoleEl|al|alsas|3)|°% | ep leal Ss a8] S/sl]le]s] Sie! & § |e 1 | 4 (58) @ Sq} QTR SIS] EB S)alSiss) gg] 8] Ss] a a4) 2/3) 2 S/R) Sle] & lee] c/a) skal & 2 lolawl4 /dj/2id] si sldolea |Hls)a le |< Campanula uniflora L.......... boo ee Nee 2) Ue * * i Aster sibiricus L............2. Bail ose alle nnekenadlling eal aattallesors etal dea) cecal! “% * Erigeron uniflorus L........... PE a a FE) RE * * # eh e E. ‘alpinus Tats sre at ehuiete pa atyats hes red said (a +] #) e) e at (lag E. grandiflorus Hook.......... = E. compositus Pursh.......... i anes ® Artennaria alpina (L.) R. Br..| *.} * | # «| wy 8 * i A. candida Greene............. ‘ eee tae ees Achillea borealis Bong..... Syne [re eel ease] Sea lee sales gal eeealeusllesselvca ulna cl eae nscale gaaleowalt fe gens Matricaria inodora L. v. grandi- flora (Hook.) Ostf........... * 7 - ba Ec cealal ere * mn eee ree Chrysanthemum integrifolium RUE Insecta Wale see ae eeeneee eine * a * C. arcticum L..........6. 02s bee eaaliee sal al lla * * * Aviemisa Rivordsomans: Bese do ov alleoe cbvaarlss vol ees eles v] daeleeeele ge afpawc]ecsals cease oclaawc) segue ose A. vulgaris L.v. Tilesii Ledeb.|....]....]....) * | * | *] * * rade * Petasites frigida (L.) Fr........ * ip ey PR el oe * aol Arnica alpina (L.) Olin........ a] =| el *| # * * Senecio palustris (L.) Hook....| * id (Deal Tees * dll a S. resedifolius Less............. cetaliaarl @ealeexal *P Pi * 7 Pies S. frigidus Less...........-...+ * a all * » sae eee Saussurea angustifolia DC...... ee * an Cee eee * Taraxacum lyratum (Ledeb.) Ciruspatnerusient staal aneleas sce T. ceratophorum (Ledeb.) DC.| *| * * A ifessees nied * * Crepis nana Richards......... 76 B Canadian Arctic Expedition, 1913-18 With respect to the vegetation covering the arctic coast, explored by the expedition, the arctic element is by far the most predominant. It is}indeed readily to be seen that of the 230 species collected not less than 84 are circum- polar; these are as follows: Cystopteris fragilis Equisetum arvense Lycopodium Selago Hierochloé alpina Alopecurus alpinus Arctagrostis latifolia Trisetum spicatum Poa arctica Arctophila effusa Dupontia Fischert Festuca ovina, mostly the var. brevifolia Festuca rubra var. arenaria Eriophorum Scheuchzert os vaginatum angustifolium Carex ancurva rigida subspathacea rariflora misandra pulla Juncus biglumis Luzula hyperborea “spicata Tofieldia palustris Saliz glauca “reticulata polaris Polygonum viniparum Oxyria digyna Lychnis apetala “ affinis Stellaria humifusa Cerastium alpinum Halianthus peploides Alsine verna var. rubella Caltha palustris Ranunculus hyperboreus oe pygmacus lapponicus nivalis Papaver nudicaule “ “cc “cc (<3 “cc “cc “ “ Draba alpina «nivalis fladmzensis hirta Eutrema Edwardsit Cardamine pratensis Chrysosplenium tetrandrum Saxifraga rivularis cernua ne Hirculus “nivalis a hieraciifolia decipiens oppositifolia Sedum Rhodiola Dryas octopetala Potentilla palustris “< nivea Rubus Chamaemorus Astragalus alpinus Phaca frigida Empetrum nigrum Epilobium latifolium « angustifolium Hippuris vulgaris Ledum palustre Loiseleuria procumbens Cassiope tetragona Arctostaphylos alpina Vaccinium uliginosum v. micro- phyllum ry Vaccinium Vitis-Idaea Polemonium boreale Myposotis silvatica Pedicularis lapponica hirsuta sudetica Pinguicula vulgaris . Erigeron uniflorus a alpinus Antennaria alpina Artemisia vulgaris v. Tilesti Petasites frigida cc (T3 “cc The Gramineae and Cyperaceae are the best represented, each by 9 species; then follow the Caryophyllaceae and Saxifragaceae each with 7, the Cruciferae with 6, the Compositae with only 5, the Rosaceae with 4, ete. By comparing the species found i in the region explored by the expedition with those reported from Spitzbergen by Nathorst, Andersson and Hesselman (l.c.), it will be seen that 84 species are common to both, and of these 59 are among the circumpolar ones. On the other hand the flora of Spitzbergen con- tains 40 species which are absent from the region visited by the expedition: these species are as follows: Arctic Plants *“*Hquisetum scirpoides : Geographical Distribution * Koenigia islandica 77 B “ tenellum *Arenaria ciliata wp om pratensis *Alsine Rossiz alpina * « — diflora P “stricta **Sagina nivalis jGlyceria ee ** Ranunculus glacialis ahliana ** Draba arctica a _ Kjellmanni * « — oblongata vatabrosa concinna « altaica ue algida *Cochlearia fenestrata aq tita caespitosa *Arabis alpina ite alpina *Cardamine bellidifolia Calamagrostis stricta Potentilla fragiformis {Carex ursina ks “ maculata “ lagopina “« multifida oe «« glareosa *Sazifraga stellaris forma comosa “ parallela *Cassiope hypnoides * “ nardina **Gentiana tenella Salix polaris **Paraxacum officinale *B t i * 6c etula nana phymatocarpum However, of these 40 species, 22 are recorded by Simmons from the arctic American archipelago (l.c.), and I have marked these with an asterisk in the list, given above. In other words, of the 124 species known from Spitzbergen, 106 occur also on the arctic coasts of this continent. And if we extend the com- parison to the coasts of Greenland, 11 species may be added (marked with twc asterisks) which the northern region from the east coast of Greenland to Point Barrow has in common with Spitzbergen. The species which thus inhabit Spitzbergen, but which are absent from this part of arctic America, including Greenland, are as follows: Equisetum tenellum Poa stricta Catabrosa concinna Saliz polaris Draba altaica Potentilla fragiformis sa multifida Equisetum tenellum (Liljebl.) Krok was first described by Liljeblad,} who referred it to E. hiemale as a variety, but with the remark that it might represent a distinct species; it was found in Torne Lapmark, and the mountains of Jimtland. Blytt (Norges Flora l.c. p. 6) refers it to £. scirpoides Michx.; the variety tenellum of E. palustre described by Elias Fries? is, of course, a different lant. : Poa stricta Lindeb. is known also from some mountains in Sweden and Norway, but with the exception of its occurrence in Nova Zembla, no stations in Russia or in Siberia are recorded by Trautvetter.’ Catabrosa concinna Th. Fr. has also been found in Nova Zembla, Waigatsch, and some few stations in western arctic Siberia: Jalmal and Dickson’s harbour. Salix polaris Wahlenb., on the other hand, is widely distributed from Dovre, Finmark, Lapmark, through arctic Russia and Siberia. From this continent, _but outside the region visited by the expedition, Salix polaris has been reported from Port Clarence and from St. Lawrence island by Kjellman.‘ Draba altaica is a native of the Altai mountains, but it has furthermore been reported from arctic Siberia (Dickson’s harbour), and from Nova Zembla. Potentilla fragiformis Willd. has been collected on Nova Zembla, in arctic 1 Utkast til en Svensk Flora. 2nd. Edit. Upsala, 1793, p. 384. 2 Flora Scanica. Upsala, 1835, p. 155. 3 Inciementa Fl. Phaenog. Rossicae (l.c.). 4 Fanerogamer fran Vest-Eskimfernas land (Vega Exped. Vetensk. Iaktt. Vol. 2. Stockholm, 1883.) Fanerogamfloran p& St. Lawrence én. (ibidem, Vol. 2. Stockholm, 1883.) 78 B Canadian Arctic Expedition, 1913-18 Russia, according to Trautvetter (I.c.), beside at some few stations in arctic Siberia. Potentilla multifida L. is very rare in the Alps of Switzerland, but has also been reported from Caucasus, southern Russia, and Siberia. With regard to Greenland the table shows that 129 species of the Greenland flora occur also in the region explored by the expedition; of these species 76 are among the circumpolar ones. Common to Greenland and the islands of the arctic American archipelago, but absent from our region are: Woodsia tlvensis (L.) R. Br. W. glabella R. Br. Lycopodium annotinum L. Agrostis canina L. A. borealis Hartm. Calamagrostis lapponica (Wahlenb.) Hartm. Aira caespitosa L. _ A. atropurpurea Wahlenb. Catabrosa aquatica (L.) Beauv. Pleuropogon Sabinii R. Br. Poa pratensis L. ‘ P. alpina L. P. laza Hnke.? Glyceria Vahliana Liebm. Agropyrum violaceum (Hornem.) Lge. Elymus arenarius L. Kobresia caricina Willd. Carex nardina Fr. . ursina Dew. . gynocrates Wormskj. . lagopina Wahlenb. . glareosa Wahlenb. . alpina Sw. . pedata Wahlenb. . ustulata Wahlenb. . capillarts L. Juncus castaneus Sm. RARQRARAARA”R Luzula parviflora (Ehrh.) Desv. Salix herbacea L. Betula nana L. Koenigia islandica L. Sagina nivalis Fr. Alsine biflora (L.) Wahlenb. Arenarta ciliata L. - Ranunculus glacialis L. (?) Cardamine bellidifolia L. Arabis arenicola (Richards.) Gelert. Arabts alpina L. Saxifraga Atzoon Jacq. S. stellaris L. Potentilla maculata Pourr. P. tridentata Soland. Pyrola minor L. Phyllodoce caerulea (L.) Gren. et Godr. Cassiope hypnoides.(L.) Don. Arctostaphylos Uva-ursi (L.) Spreng. Diapensia lapponica L. Euphrasia latifolia Pursh. Pedicularis flammea L. Campanula rotundifolia L. Gnaphalium norvegicum Gunn. Artemisia borealis Pall. Taraxacum phymatocarpum Vahl. in all, 53 species, 23 of which are Gramineae and Cyperaceae. Among these species, which the arctic American archipelago has in com- Agrostis canina Aira cespitosa Catabrosa algida Poa alpina Agropyrum violaceum Elymus arenarius Carex gynocrates C. glareosa C. alpina C. ustulata Juncus castaneus Luzula parviflora. Salix herbacea Betula nana Koenigia islandica in all, 30 species. mon with Greenland, the following occur furthermore in the Hudson bay region; Sagina nivalis Cardamine bellidifolia Arabis alpina Sazifraga Aizoon S. stellaris Potentilla maculata P. tridentata Pyrola minor Phyllodoce caerulea Casstope hypnoides Arctostaphylos Uva-ursi Diapensia lapponica Pedicularis flammea Campanula rotundifolia Artemisia borealis Arctic Plants: Geographical Distribution 79 B Of these species the following are circumpolar: Aira caespitosa Cardamine bellidifolia Catabrosa algida Arabis alpina Carex glareosa Saxifraga stellaris Luzula parviflora Campanula rotundifolia Salix herbacea Artemisia borealis _ However, the Hudson bay region, between 55° and 65° N. L., is quite rich in arctic species, and of the about 450 species which are known from there, about 90 are circumpolar, including those enumerated above. Most of these circumpolar plants evidently belonged to a flora, which in tertiary times inhabited the polar regions, but was forced towards the south during the glacial epoch. Many of these sought refuge in the higher mountains, but returned to their northern homes when the ice receded. And on their retreat to the polar regions they were accompanied by a number of southern, mainly alpine, species; thus the present arctic, flora is also represented by a southern element. However, the probability is that not a few of the circumpolar plants remained on the southern mountains, some to persist, others to succumb. This may be demonstrated by the indisputable fact that the alpine flora shows a commingling of types similar to the present arctic flora. These foreign arctic types are remarkably distinct, according to longitude, an American element being represented in the arctic American flora, a European in the arctic European, and a Siberian in the arctic Asiatic. For instance, the occurrence of the American Astragalus aboriginorum in arctic America, the American Erigeron compositus in arctic America and Greenland, etc. But it is merely conjectural to determine the actual roads, followed by these plants from the polar regions to the south, as well as on their retreat. And the difficulty culminates especially when we bear in mind the extraordinarily scattered distribution of some of these polar plants in southern regions, the higher mountains for instance. It is still more difficult to appreciate the actual means by which the migration became effected. Great importance has been attributed to the belief that migratory birds carry with them seeds of plants, but recently we have learned that such birds migrate on an empty stomach, and are almost always clean when they commence their long journey. This statement we owe to Knud Andersen,’ who made the start- ling observation in Denmark, that during a period of 4-5 years the intestines of all the birds killed near the lighthouses were found to be empty, beside the fact that no seeds were found adhering to any parts of these birds. However, Pro- fessor Henry W. Henshaw? has, more recently, expressed an opinion which is entirely different, making the following statement: “The sources of the vegetation and the means by which the seeds of plants and shrubs were originally transported to these distant ocean-girt islands, thousands of miles from the nearest mainland, are a most inviting field of specula- tion. The winds are capable of conveying minute seeds to great distances, and favourable ocean currents also materially aid as plant distributors. Birds, however, are doubtless the most important of nature’s seed carriers. Viscid ‘and hooked seeds attach to their plumage, or seeds may be carried in smears of earth or mud on feathers, bill, or feet. Such seeds may be transported indefinite distances and, once in a thousand years or so, dropped on soil favourable to growth. The members of the most recent expedition to the island, the Nutting party in 1911, were especially requested to examine carefully all their specimens of Laysan birds for the presence of seeds, and actually found attached to the foot of a Laysan albatross a seed of a species of the bean-caper family, which is generally distributed in the South Sea islands. So far as known the plant does not grow on Laysan, and had this seed chanced to fall in a favourable spot the ._H. Ostenfeld in Botany of the Faeroes. Part I, p. 117. Copenhagen, 1901. : concen) W. Our Mid-Pacific Bird Reservation. Yearbook of the Dept. of Agric. 1911. Wash- ington, D.C. 80 B Canadian Arctic Expedition, 1913-18 flora of the island might have been enriched by another species. A still more striking instance of a bird acting as a carrier of seeds was observed several years ago by Mr. Bryan on Marcus Island, where he found no fewer than 40 seeds of a tree of the madder family adhering to the feathers of a shearwater. These and other similar facts show how the flora of oceanic islands may be transported by birds from island to island or from continents to distant shores.” Furthermore, that resident birds, as for instance the partridge, may well transport many seeds from one place to another, has been pointed out by Dar- win.! 7 With regard to other factors of dispersion, water for instance, we know, so far, too little about the possibility of seeds retaining their vitality during pro- longed immersion in salt water. The transport by means of ice may take place, as pointed out by Simmons, when the question merely deals with narrow chan- nels or fjords, but, nevertheless, more indirectly so, since the wind becomes a more or less necessary, additional factor for carrying the seeds to the shore. However, even if we have reached no decisive conclusion as to the means by which the migration was effected, nor by what roads it took place, we do know that the plants exist there, that arctic and alpine species do occur in the polar regions, and quite a number of these furthermore in the mountains farther south. It seems even plausible to suppose that the mountains harbour a larger number of arctic remnants than has really been accounted for, because the arctic species in the mountains as wel! as the alpine species in the arctic region may have submitted to a more or less pronounced modification in structure due to the change of conditions. Such modification, in external structure especially, may, as stated above, have resulted in the respective species appearing in dis- guise, and being mistaken for separate species. In this way botanical manuals, comprising floras of mountainous regions, are sometimes crowded with sup- posed new species which actually are simply modified forms or varieties produced by change of environment, especially climate and soil. Consequently, when dealing with plants scattered over large areas, we must consider whether they represent analogous types or mere forms. Artemisia, Antennaria, Draba, and several other genera have recently furnished an almost’ untold number of sup- posed new species, regardless of the natural modification to be sought in ‘‘geo- graphical distribution.” Returning to the circumpolar species, these having been so admirably des- cribed and their distribution discussed by some of the ablest botanists, the danger of errors in their determination is but slight, even if more recently some authors have ventured to make changes—as a rule wholly unwarranted. Having enumerated the circumpblar species, collected by the expedition, we shall now offer a few data regarding their occurrence in other regions; more- over, some points will be discussed relative to the southern, the alpine element, which has also become distributed in the polar regions. In accordance with the enumeration of the circumpolar species, given in the preceding pages, and the accompanying table, it will be seen that those exhibiting the widest geographical distribution farther south are as follows: Cystopteris fragilis, Equisetum arvense, Lycopodium Selago, Trisetum spicatum, Poa arctica, Festuca ovina, especially the variety brevifolia, F. rubra var. arenaria, Eriophorum Scheuchzeri, E. vaginatum, E. angustifolium, Carex incurva, the distri- bution of which on this continent extends as far south as the Strait of Magellan, C. rigida, which, however, in the Caucasus occurs only as the variety dacica, a plant, which Boissier (FI. orient., Genéve, 1867-84) has enumerated as C. rigida, but by Richter (Pl. Europ., Leipzig, 1890-1903) has been referred to C. caespitosa v. dacica. Furthermore Carex misandra, C. pulla, Juncus biglumis, Salix reticulata, Polygonum viviparum, Oxyria digyna, Lychnis apetala, Cerastium 1 Darwin, C. On the Origin of Species. Fourth Ed. London, 1866, p. 432. See also: Hesselman, H. N&gra iakttagelser fver vaxternas spridning. Botan. Notiser, 1897, p. 97. Arctic Plants: Geographical Distribution 81 B alpinum, Papaver.nudicaule, Draba fladnizensis, Cardamine pratensis, Saxifraga cernua, S. Hirculus, S. nivalis, S. hieraciifolia, S. decipiens, S. oppositifolia, Sedum Rhodiola, Dryas, Potentilla nivea, P. palustris, Empetrum, Epilobium, Erigeron uniflorus, and E. alpinus. From this small enumeration it will be seen that most of these species may have their spores, seeds, or fruit readily dispersed by means of the wind, especially the Pteridophyta, the Gramineae, Cyperaceae, notably Eriophorum, the Saliz, Oxyria, Dryas, etc. But Sazxifraga cernua, Polygonum viviparum, and perhaps also Cardamine pratensis form an exception, since the production of bulblets has taken the place of fertile flowers; with regard to Empetrum, the seeds of the fleshy fruits may be dispersed only by means of animals. However, many of the other species are just as well adapted to be dispersed, especially by the wind; thus their more or less scant distribution farther south may be due to their local extermination, or perhaps because they may represent young species of northern origin. I think especially of the willows, which un- doubtedly have an important centre in the northern regions. | With regard to the truly arctic element, there are several species which, although not being circumpolar, may be looked upon as genuine arctic types, and some of these have not, so far, become distributed farther south. Among these may be mentioned Hierochloé pauciflora, Poa abbreviata, Glyceria, Carex stans, C. subspathacea, C. reducta, C. compacta, Luzula nivalis, Salix rotundifolia, and several other species of Salix, Rumex, Lychnis Taylorae, Alsine macrocarpa, Merckia, Anemone Richardsonii, Ranunculus Pallasii, R. Sabinit, Lesquerella, Cochlearia, Draba corymbosa, Braya purpurascens, Hesperis, Cardamine digitata, Parrya arctica, Saxifraga radiata, Dryas integrifolia, Sieversia, Potentilla pul- chella and several other Potentilla species, Rosa acicularis, Lupinus, Hedysarum, Selinum, Pyrola, Cassiope, Dodecatheon, Primula borealis, P. stricta, Gentiana, Phlox, Meriensia Drummondii, Pedicularis hirsuta, P. arctica, P. capitata, Lagotis, Aster, Achillea, Chrysanthemum integrifolium, Artemisia vulgaris var. Tilesit, Arnica, Senecio resedifolius, Saussurea, Taraxacum, and Crepis. About 60 species, and adding to these the 80 circumpolar ones, we have in all 140 species, collected by the expedition, which may be termed strictly arctic, so far as we know, at least in accordance with their present distribution, being absent from regions farther south. The remaining species, about 90, may be partly such as have originated in the arctic regions, but of which a certain element remained in the mountains, while the others returned. And another part consists of southern species, which accompanied the arctic on their retreat. These southern species cannot with certainty be defined, except that their distribution in the arctic region, being either very scant or scattered, might indicate their southern origin. As far as concerns the structure of arctic species, it has been shown in the preceding pages that no morphological structure seems absolutely characteris- tic of these; they share the same development of their floral and vegetative organs with their southern, allied species; they exhibit exactly the same method of vegetative reproduction as these, and are, indeed, in many cases not of such dwarfed stature as frequently described. But they resemble alpine plants, and they are, with but very few exceptions, perennial like them. There is thus no absolute morphological character by which arctic and alpine species may be defined, and we must therefore consider them from other points of view, when the question arises to distinguish them, in the present case to draw the line between the northern and southern element. It may be possible to establish such a distinction by means of comparing the genera, or in some cases the species. And in this case we are allowed to include, but only as supple- mentary, peculiarities in their morphological structure. ; In the subsequent pages I shall endeavour to show the importance of the present geographical distribution in defining the probable location of the 24657—6 82 B Canadian Arctic Expedition, 1913-18 centre of these species, their centre of development and of distribution. For this purpose I have especially selected such species as are not circumpolar, and of which the distribution in the south is well exemplified. In beginning with the Pteridophyta, the species collected by the expedition are too few to give us any idea of the representation of this group of plants in the arctic region. Therefore, in the accompanying table (Table 2), I have enumerated all the species known from our arctic shore, including the islands of the archipelago, and from Greenland; however, in view of the fact that several species reported from Greenland have not, so far, been found on our arctic shore, a special column has been given to the species known from that country. On the other hand, no column has been given to arctic Siberia, since, according to Ledebour (FI. Rossica l.c.), Dryopteris fragrans is the only fern recorded from © ; a3 : S yg . TABLE 2, ‘Sg a ‘ $s F g < Geographical distribution of arctic Pteridophyta, By g 3 g ae ne Q & ee ine 85/2] 8/0 | 8 lasl z 2/8,|4 21s /#|4/2le|<|5/ se" Phegopteris polypodioides Fée..... .. ....... 0.0.0... saga | cians pe Pelee ec bial ea P..Dryopiteris (lj.) POs scsi canis d connireprandavereec: io had ee a cd Dal a] eee Pe Cryptogramma acrostichoides R. Br.................... * oe ene hi ae ee Asplenium viride Huds.............0.0.000000 cece eee cll ius Delf #] oe Polystichum Lonchitis (L.) Roth........... Ls cues a ee ee ee Dryopteris fragrans (L.) Schott................. 00.0, i | ees eae * » 2 D. spinulosa (Mill.) Kze............000 00.0.0 ween ee im a ee oe ee Blechnum spicant (L.) Roth.............0.0.0. cece eee bass aaah @ Meine Cal x onal lead Cystopteris fragilis (L.) Bernh........................ kc iV | co W oodsia ilvensis (L.) R. Br..........0 20.0200 eee Re eect A) EL A A a ae W. hyperborea R. Br...........0.000000 0c ee eee cal (imac (eer Rill OL (Rg eee) (eee eee MA W. glabella R. Breoceeeccccee cece ceeeeeeeeeee veeees ee) eel Neclaalacd #2 Botrychium Lunaria (L.) Sw.........0...0.... ree Wee aca ere (| | B. lanceolatum (Gmel.) Ang..................: roe ie Toca bce oF Hquiseline ar vense Veena 304405468 he ba be Lwnsn ih daaean * * * * * * * * * * * GOST TAL Oy ena een eee nn dessa F [empganl) EA: = BE Pat wl +l #] | « Pie POUUSEPE, Wiss ssi cisacimrondyasdosaceer ch uanoritaita Sothern Waansbatoieen a | «| a) «lel ele] el x) Drs Onno nbeiny Wisi syajesissca sad naa cscabaon cand waveania auttvecen ie * Sz * * * * * TS BUD VINE TS 5 srt 2 cascaseeatase eae nla taolat staple naw deeatbie beg cal |e # eae es OE cat OME ae Selaginella selaginoides (L.) Lk..........0... 000000000. | pel «| oe] ee oe Isoétes echinospora Dul....... 00.6.0 00 ccc cece cece eae Lect © leeuel eae Fes | can | ae eae Arctic Plants: Geographical Distribution 83 B that region.: “Ad lacum Falkadja Lat. 73° 30’ prope flumen Taimyr rarissime (Middendorff et Ruprecht).” Besides this fern, Lycopodium Selago, L. complan- atum, and Equisetum arvense are the only Pteridophyta recorded from arctic Siberia, and by adding their occurrence in Siberia to the other stations given in the table (Table 2) it will be seen that Lycopodium Selago and Equisetum arvense are circumpolar, in fact the only circumpolar Pteridophyta known so far. Lyco- podium complanatum, on the other hand, is not circumpolar, since it is absent from arctic America. _, With regard to the occurrence of the Pteridophyta in the arctic region, I wish to offer some few statements relative to Greenland. Between Lat. 62° and 72°, Frederikshaab-Upernivik on the west coast, ferns are not uncommon; they are especially well represented in crevices of rocks near mountain brooks, for instance, Polystichum Lonchitis, Phegopteris polypodioides, P. Dryopteris, and Dryopteris spinulosa; on dry rocks we meet with the species of Woodsia, Dryopteris fragrans, and Cystopteris. The very rare Asplenium viride I found growing in deep shade, in a grotto, associated with luxuriant specimens of Dryopteris. The Equisetaceae, notably E. scirpoides, are quite common, and among the Lycopodiaceae, L. alpinum and L. Selago are very frequently met with, and sometimes quite abundant. The latitude reached by some of these species in Greenland is quite remarkable; the following quotations may be of interest: Cystopteris fragilis 76° Woodsia hyperborea 73° 21’ Dryopteris fragrans 72° 53’ (73° 30’ in Siberia) Woodsia ilvensis 72° 48’ Equisetum arvense 72° Woodsia glabella 71° 67’ Lycopodium Selago and Equisetum variegatum 71° Equisetum scirpoides, E. silvaticum, and Lycopodium alpinum 70° Botrychium Lunaria 69° 20’ Phegopteris Dryopteris and Polystichum Lonchitis 69° 15’ Phegopteris polypodioides and Dryopteris spinulosa 65° 40’ Selaginella selaginoides 64° 15’ This northern distribution may be extended farther if we compare the _stations, known from Spitzbergen, as recorded ‘by Nathorst (l.c.), Gunnar Andersson and Hesselman (l.c.); according to these authors the following latitudes may be quoted: Lycopodium Selago and Equisetum variegatum 80° Cystopteris fragilis 79° 15’ Woodsia glabella 78° 40’ Equisetum arvense 78° 30’ E. scirpoides 77° 30’ With regard to the distribution farther south, we see from the table (Table 2), that with the exception of Lycopodium alpinum and Isoétes all the other Pteridophyta have been found on this continent, south of the arctic circle, and notably in the mountains. Cryptogramma, Botrychium lanceolatum, Asplenium viride, Isoétes, and Dryopteris fragrans, having been recorded from very few and scattered stations, may be treated separately. adese. eee ee With regard to Cryptogramma acrostichoides, the distribution of this species extends from lake Huron westward to British Columbia, stretching northward to within the arctic circle (Macoun I.c.), and according to Gray (1.c.), it is known also from Colorado and California. Another species, C. Stellert (Gmel.) Prantl (Allosurus Stelleri Rupr.), occurs also on this continent from Labrador to British Columbia, south to Illinois and northern Pennsylvania; it occurs 24657—64 84 B Canadian Arctic Expedition, 1913-18 also in eastern Siberia and in the Baikal mountains. Finally, if Pellaea densa Hook. is also to be referred to this genus, we have a third species distributed from Quebec to British Columbia. The geographical centre of the genus appears thus to have been located on this continent, but south of the arctic regions. Typical Botrychium lanceolatum is a native of Greenland (61° N. L.) and arctic Scandinavia, and also of Dovre, Herjedal, etc., but is on our continent repre- sented by a variety angustisegmentum Pease and Moore. It is evidently, of Scandinavian origin. The same appears also to be the case of Asplenium viride, and its present southern distribution may indicate that it is one of the northern plants which were left over on the southern mountains while the others retreated to their northern homes. With regard to Iscétes echinospora, its present distribution: Iceland, Faeroe Islands, Scandinavia, Denmark, and Central Europe, seems to indicate that the geographical centre of this species is located in Central Europe, and that the distribution to the north, to Greenland and Finmark, took place during the glacial epoch. Dryopteris fragrans is, on the other hand, a genuine arctic type, which was evidently more widely distributed in the polar regions before the glacial epoch, and its occurrence in arctic Siberia, Ural, and Altai, and also in Kamtchatka, might indicate a former, circumpolar distribution. Before we can consider the distribution in general of all the other Pteri- dophyta, enumerated in Table 2, we might add the following ten species which have been found in the arctic regions of Scandinavia and Russia, but not in arctic America; by adding these, we shall have, I believe, an approximately complete list of all the arctic Pteridophyta. The number of species is: Filices, 24; Equise- taceae, 5; Lycopodiaceae, 3; Selaginellaceae, 1; and Isoétaceae, 1. vo G C o 8 g 3 2 5 TABLE 2A 3 a 2 = : 5 3 | 38 8 I a Geographical distribution of arctic Pteridcphyta, zg B Ben c fQ 4 absent from arctic America. 8 3 & a s 3 wn fot Bo Z 3 E s 2 | 2 }ee] s a | 3 e S Fe) o a oa Rs 2 |e bee |e) 2 } 8] & | es < a < ) ‘S < Polypodium vulgare Vie: wissen ca saciewr wea sae ia sae * ” = * i ~ * PPG Toss wacemcte diane diestemrandayagieweds ig Bae oe eee BM a aie cateaall ey geese Asplenium Filit femina (L.) Bernh............ ees * * * * * + a leece Al, SONU IONOLE BW, oc occa ccecnnsabnesasawserenres ae ree * * * % ame foeene PA SET GYUGUM TUT ENT I eyeing one aren eiasond ravi rable anny el lunges 4 iNedevtan onl baceenavennll avieavecerst| untidecees * * Cystopteris montana Bernh................0..0000005 aoe * aa ere ee ee er ee id Allosuus crisps Bettis rey sccsweecndveins 2eeas . Bessy. | eed octet ueealeaisacllonanex Dryopteris Filiz mas (L.) Schott................... ‘.) . * * * * a Cte Onoclea Struthiopteris (L.) Hoffm.............. 0 0... Je..00. * = x * * * * Botrychium matricarioides (L.) Willd........ .......J...... ¥ 3 © les ee cel epareelegaxerlaeneus As may be seen from these tables, most of the species have also been reported from the mountains farther south; not less than 29 from America, 25 from Altai, 24 from the Alps and Pyrenees, 18 from Ural and Kamtchatka, and 16 from Caucasus. Among these, the five species which are absent from America are: Arctic Plants: Geographical Distribution 85 B Polypodium rhaeticum, Asplenium crenatum, Allosurus, Lycopodium alpinum, and Isoétes. Absent from the European Alps, but recorded from Altai are: Dryopteris fragrans, Asplenium crenatum, Woodsia glabella, W. hyperborea, Equisetum scirpoides, andE. variegatum. Finally, absent from Altai but reported from the European Alps are: Cystopteris montana, Allosurus, Aspleniwm viride, Blechnum, and Botrychium matricartotdes. It is interesting to notice that, if those most widely distributed had also been reported from arctic Siberia, 17 of the species enumerated on Table 2 would be circumpolar. Their total absence from that region makes a wide gap in their northern distribution. And naturally so, for the Siberian tundra would never make a home suitable to ferns. Nevertheless, on the southern border, in the regions of Altai and Baikal, we have seen that not less than 25 species of arctic Pteridophyta have become established, and strange as it might appear, associated with flowering plants, many of which are known, at present, to occur in arctic Siberia, and even to be circumpolar. Otherwise with Europe and North America. We have seen that the distri- bution of the arctic Pteridophyta, in the temperate regions, notably mountainous, of these continents corresponds well with their northern occurrence. And at the same time we have seen that several of the species are identical with those of the Altai mountains. In other words, there is a striking accordance between the flora of the southern mountains throughout the northern hemisphere and that of the arctic region, as far as the Pteridophytaare concerned. And wehave seen also that in the polar regions the ferns and their allies have reached the highest latitudes known for vascular plants. By combining these data relative to their present distribution, I really believe, that most of the arctic Pteridophyta originated in the far north. They cannot possibly be considered as a part of the original alpine Altai flora, as long as we have no trace of their migration from these mountains to the north, such as is otherwise the case of a number of flowering plants, still to be found north of Altai, i.e. in arctic Siberia. In passing to describe the geographical distribution of the Phanerogams, it will be seen that we are dealing with a group of plants which has become more evenly distributed around the polar regions than the Pteridophyta; thus a number of them are, at present, circumpolar. Although Picea canadensis BSP. (Abies canadensis Mill.) does not proper- ly belong to the arctic zone, it deserves mention that it is known to occur in many places far north of the arctic circle in Canada and Alaska. It was found in abundance in two districts explored by the expedition, namely the Mackenzie delta and Coppermine river valley. With regard to the distribution of the species in the Mackenzie delta, Mr. A. H. Harrison has published a map of the delta showing the north limit of the spruce.!_ According to this map the spruce begins a little above Lat. 69° N., at the south point of Richard island, extending from there southeastward to second Eskimo lake, at about Lat. 68° 50’ N., and then northeastward, reaching Lat. 69° 35’ N. near the south shore of Liverpool bay. The spruce thus comes within ten miles of the coast on the west side of Franklin bay, on the Horton river, and within thirty or forty miles of the coast on the Anderson river, south of Liverpool bay. The most northern bunch of spruce trees in the Coppermine river region is, according to Dr. R. M. Anderson (in Stefansson: My Life with the Eskimo, l.c.p. 445), within six miles of the coast on a little creek valley several miles east of the Coppermine river, but these trees were scrubby and dwarfed. In a small, isolated grove of spruce near Kendall river, a few miles west of the Coppermine river, and not far from Dismal lake, Dr. Anderson measured one tree, which was four feet and six inches in circumference five feet above the ground and above the bench roots; the same 1 “In Search of a Polar Continent.’’ 1905-1907. London, 1908. 86 B Canadian Arctic Expedition, 1913-18 tree was five feet in circumference three feet above the ground. Dr. Anderson furthermore states that none of these trees were very high, but were mostly straight-grained, and not twisted spirally as are most of the spruce in this region. These northern groves of spruce have been mentioned already by the earliest expeditions which visited the lower Coppermine river, notably by Samuel Hearne (1769-1772), Captain John Franklin (1819-1822), and Dr. John Richard- son (1825-1827). In his ‘Arctic Searching-Expedition, 1848-1849,’ Dr. John, Richardson gives an interesting description of the spruce, as he found it above “Bloody Falls’ on the Coppermine river, which may be quoted as follows: “(Crossing overland from west of the mouth of Coppermine river to above Bloody Falls) we encamped on the 7th September on coming to a clump of stunted white spruce . . . . In the existence of many scattered stumps of decayed spruce fir trees, and the total absence of young plants one might be led to infer that of late years the climate had deteriorated and that the country was no longer capable of supporting trees so near the sea coast as it had formerly done. The largest trees in the clump in which we bivouacked had a circumfer- ence of 37 inches at the height of four feet from the ground. Its annual layers were very numerous and fine, and indicated centuries of growth, but I was unable to reckon them. This place lies in Lat. 67° 22’ N. . . . . One circum- stance which came under my observation, and has been cursorily alluded to, is the existence of very ancient stumps of trees, either solitarily or grouped in various places of the barren grounds, seemingly the vestiges of the forest which had spread more widely over the country some centuries ago than in the present day. . . . On the sheltered banks of rivers, even in the barren grounds, clumps of living trees occasionally occur, but the stumps I speak of stand often on the exposed side of a hill, and indicate a deterioration of the climate, however that may have been produced. We saw no young firs growing up in such situa- tions to leave similar vestiges in a future age. . . . . Within the Arctic circle it (the white spruce) seldom exceeds 40 or 50 feet in height, though in ravines where it is well sheltered, and has a suitable soil it attains twice that altitude. Its age in these high latitudes exceeds 400 years before it shows signs of decay.” Similar observations were made by Mr. Frits Johansen, who has published an interesting arcticle on “The Forest’s Losing Fight in Arctic Canada,”! in which the author calls attention to the scarcity of very young spruce trees in this northern limit of their growth. Mr. Johansen attributes this to the intense cold and the sweeping winds in the winter killing off most of the small seedlings which may have developed during the summer. This explanation seems well founded by the fact that even the small and stunted trees were found to be about half a century old, while the largest trees would reach an age of almost 500 years. Furthermore, Mr. Johansen made the interesting observation that forest insects are principally responsible for the killing of the trees or tree parts in the lower Coppermine river, a fact which has, so far, escaped the attention of the various explorers who have visited the region from time to time. Asa matter of fact, the very isolation and exposure of the individual trees here makes them an easy mark for attacks by forest insects, which by living under the bark are less influenced by the shortness of the summer season. : As to the occurrence of spruce on the arctic side of the mountains west of the Mackenzie delta, Dr. R. M. Anderson writes: ‘Information which I have obtained from natives and reliable white hunters and prospectors agrees that west of the Mackenzie delta there are spruce trees on Babbage river (back of Kay point, Y.T.), and on Firth river (the so-called ‘Herschel Island river,’ on both sides of the International Boundary), but none on rivers west of these. 1 Canadian Foresty Journal. Ottawa, July, 1919, p. 303. Arctic Plants: Geographical Distribution. 87 B Some Colville river Eskimos told us that there are a few spruce on the north side of the mountains near the head of the Itkillik river, a large tributary of the Colville river on its eastern side. No other rivers on the north slope of Endicott range have coniferous trees.” Fig. 1. Picea canadensis and Saliz on creek slope, tributary to Coppermine river below Sandstone rapids. gs Feb. 15th, 1915. (Photo by F. Johansen.) : ; Fig. 2. Picea canadensis in creek bed, tributary to Coppermine river, below Sandstone rapids. February 15th, 1915. (Photo by F. Johansen.) : 88 B Canadian Arctic Expedition, 1918-18 Fig. 3. Grove of white spruce (Picea canadensis) southeast of Sandstone rapid, Coppermine river. The most northern dense grove in this region, a thick stand being found only where protected from sweeping winds and with good exposure to the sun. February 15, 1915. (Photo by R. M. Anderson.) With respect to the general distribution of Picea canadensis, Sargent } gives this as follows: “The White Spruce inhabits the banks of streams and lakes and the borders of swamps, in rich moist alluvial soil, ocean cliffs, and less commonly at the north rocky slopes of low hills; it ranges from the shores of Ungava bay in Labrador westward to those of Hudson bay, and from the mouth of Seal river not far to the north of Cape Churchill, it is scattered along the northern frontier of the forest nearly to the shores of the Arctic Sea, and, crossing the continental divide, reaches Bering Strait in 66° 44’ north latitude. South- ward it extends down the Atlantic coast to southern Maine, growing often close to the shore, and to northern New Hampshire, Vermont, New York, Northern Michigan and Minnesota and the Black Hills of Dakota, and through the interior of Alaska and along the Rocky Mountains to northern Montana.” The accompanying photographs were taken by Dr. Anderson and Mr. Johansen and give an excellent idea of the growth of the spruce in the far north. Among the 20 species of Gramineae, enumerated on the geographical table (Table 1), ten are circumpolar: Hierochloé alpina, Alopecurus, Arctagrostis, Trisetum, Catabrosa, Poa arctica, Dupontia, Festuca ovina and F. rubra, and finally Arctophila. With the exception of Dupontia, Arctophila, and Catabrosa, all the others have also been reported from the Altai mountains; Trisetum, Poa arctica, and the two species of Festuca are known also from the Himalayas. With regard to their occurrence in the Alps and Pyrenees, only Trisetum and the two species of Festuca have been reported from these mountains. In the Rocky mountains south of the arctic region we have only four of these represented, namely, Trisetum, Catabrosa, Poa, and Festuca ovina; this seems somewhat peculiar since they all have been found on the American coast of Bering strait, and with the only exception of Dupontia and Alopecurus, they have also been reported from the Asiatic coast. However, Dupontia, Arctophila, and Arctagrostis are almost exclusively confined to the arctic regions, and the presence of Arcta- grostis in the Altai mountains may indicate that the genus reached these mountains during the glacial epoch and remained there. Besides these circum- polar species, all of which undoubtedly originated in the arctic regions, the fol- lowing may also have had their centre located in the north, viz.: Hierochloé pauci- flora, Calamagréstis, Elymus, Poa glauca, P. abbreviata, and the species of Glyceria. 1 The Silva of North America, Vol. XII. Boston, 1898. Arctic Plants: Geographical Distribution 89 B Common to both coasts of Bering strait, and furthermore recorded from Altai mountains and the Himalayas, we have thus in Festuca altaica a southern type, which evidently was one of those that accompanied the arctic plants on their retreat to the north during the glacial epoch. Hordewm jubatum may well be considered as an introduced weed. Among these arctic Gramineae two species are of special interest: Poa abbreviata and Calamagrostis purpurascens; while both are undoubtedly of arctic American origin, the former has also reached Spitzbergen, and the latter is not so very rare in the alpine region of the Rocky mountains (Colorado); besides that it occurs in Greenland, and on the American coast of Bering strait. With reference to Poa abbreviata, Nathorst. (l.c.p. 287), mentions this in connection with Glyceria angustata from Okhotsk, and Alsine Rossii from the Rocky moun- tains, as being the only species which Spitzbergen has not in common with Nova Zembla and Scandinavia. The occurrence of these three species in Spitzbergen is very strange, inasmuch as we are entitled to believe that the flora of Spitz- bergen originally came from Scandinavia and Nova Zembla in post-glacial times, when these stations were connected with each other by “land.” Calamagrostis purpurascens is one of the many American species which found their way to Greenland. Although the large family of Cyperaceae is relatively poorly represented on the arctic coast of our continent, some of the species which have been collected are, nevertheless, of some interest from a geographical viewpoint. Elyna Bellardii, for instance, shows a distribution of enormous extent, viz.: the arctic regions of America, Greenland, Scandinavia, Russia, and Siberia; farther south it is alpine in the Rocky mountains (Colorado), and it also occurs in the Alps and Pyrenees, Caucasus, Turkestan, Altai, Davuria, and Iceland. It is on this continent associated, to some extent, with a Cobresia, C. caricina Willd., which is also arctic-alpine, and known from several of the same countries, and also from the Himalayas. Of the 29 species known of these two genera, the 27 are ex- clusively Asiatic, and principally natives of the Himalayas. Elyna is thus the only one of these which may be looked upon as an arctic, circumpolar type, and it would seem very strange if the species had really originated from one single centre of distribution. No doubt the Altai mountains were an important centre for its distribution in Asia, and it might have reached the European mountains by the way of Turkestan and Caucasus. But with regard to the occurrence on this continent, and especially in the Rocky moun- tains, a second centre may have been located there, from where it thus became distributed farther north during the glacial epoch. Moreover, Elyna is, as stated above, on this continent accompanied by Cobresia caricina, and besides this, a second species, C. elachycarpa Fernald, has been detected in Maine. This species, however, is by Kiikenthal referred to the genus Carez, for no other reason than “Area geographica Cobresiae speciem excludere videtur.”’” Never- theless, there is a Claytonia in New Zealand, a Podophyllum in the Himalayas, a Jeffersonia in Manchuria, etc. Of the twelve Carices collected, six are circumpolar, and we have seen from the introduction the remarkable geographical range exhibited by Carex incurva; nevertheless, I am most inclined to, consider these circumpolar species as having originated in the north, in the polar regions; a similar northern centre may also be attributed to the non-circumpolar, but exclusively arctic C. reducta, C. stans, and C. compacta, But with reference to C. scirpoidea, this did undoubtedly originate in the Rocky mountains, where it is more amply represented than anywhere else, and accompanied by the variety stenochlaena, and in the Coast range, California, by the variety gigas. The fact that C. scirpoidea does also occur in Greenland, 90 B Canadian Arctic Expedition, 1913-18 Iceland, and Norway, seems to support the theory, proposed by Nathorst, that the American element in the European mountains may have reached these by way of Greenland during the glacial epoch. C. rupestris, not being circumpolar, but represented on our northern coast, in Greenland, Spitzbergen, Scandinavia, Nova Zembla, Iceland, the Alps and Pyrenees, Caucasus, Altai, besides the Rocky mountains, as far south as Colorado, cannot possibly have acquired such wide, but more or less disconnected range, from one single centre, located in the south, or in the north either. Carex vaginata is in the table marked down for: islands of the arctic Ameri- can archipelago, arctic Scandinavia, arctic Russia, Iceland, Alps and Pyrenees, and both coasts of Bering strait. The species is, moreover, widely distributed in the mountains of Central Europe; it occurs also in West and East Siberia, in the Amur district, and in northern Japan. Several varieties have been described, viz.: distracta Norman from Finmark, pauciflora (Wahlenb.) Ands. from East Siberia, ‘‘Lena,’’ Gruetteri Aschs. et Graeb. from the Baltic provinces, Petersii from Kamtchatka, altocaulis Dew. from Labrador to British Columbia, the northern Atlantic States, and also Michigan and Minnesota. It would thus appear as if the centre of distribution of this species may be located south of the arctic region; the occurrence of certain varieties in the south speaks also in favour of this supposition. Furthermore, Carex vaginata is a member of the Grex Cenchrocarpae,! the species of which are indeed southern types; only a very few have reached the polar regions, viz.: C. bicolor All., C. livida Willd., and C. vaginata. However, the distribution within the arctic region of these species is very limited. According to these data, the arctic Gramineae and Cyperaceae do exhibit a much wider distribution in the north than in the south, with only a very few exceptions, namely: Festuca altaica, Elyna, Carex vaginata, C. rupestris, and C. sewrpoidea. And, furthermore, there are several members of these families which, according to our present knowledge, are endemic to the arctic regions, viz.: Hierochloé pauciflora, Poa abbreviata, Arctophila, Dupontia, Glyceria, Elymus, Carex stans, C. subspathacea, C. reducta, and C. compacta; in all 12 species (Glyceria with 3), which are confined to these regions and must naturally have developed there. Arctophila and Dupontia are both well marked, and indeed, quite char- acteristic genera, and none of their species are known except from the arctic regions. Furthermore, Hierochloé pauciflora and Poa abbreviata are both excellent species, and may well be looked upon as genuine arctic types, i.e. “types” with regard to their morphological structure and geographical distribution. The species of Glyceria are also interesting from this point of view; their habit is very characteristic, when compared with that of most of the other species of the genus and, as has been shown in the chapter dealing with the mor- phological structure of the arctic plants, G. vilfoidea shows a very singular structure of the stolons, which is almost unique. But with regard to the Cyper- aceae, even if Eriophorum, at least some of the species, may have originated in the arctic, the singular structure of the fruit constitutes no character which the genus does not share with its southern representatives. And among the Carices there seems to be no type especially characteristic of the arctic regions; the genuine arctic species, not only those as are circumpolar, but also those as are confined to the polar regions, resemble their southern allies, alpine as well as lowland types. The temperate regions of both Worlds are the principal home of Juncus and Luzula; they have developed there and acquired a habit and structure which marks them wherever they occur. However, there is also an arctic element of both genera which seemingly originated in the far north and of which several types have entered the alpine regions in the south during the glacial epoch; some, but only a few, species are endemic to the arctic regions. 1 Holm, Theo. Greges Caricum. Am. Journ. of Sc., Vol. 16. New Haven, 1903, p. 458. Arctic Plants: Geographical Distribution 91 B The lowland, we might perhaps say the “maritime” and “paludine,” habit of Juncus is in striking contrast with the sylvan or frequently mountainous habit of Luzula, and wherever arctic elements of the genera meet, such as is only the case in the higher mountains, the species seemingly prove to be of high-northern, i.e. of arctic origin. : Very few of the Juncaceae are circumpolar, viz.: Juncus biglumis, Luzula spicata, L. hyperborea, and L. parviflora. Nevertheless, six species have been reported from Spitzbergen, from latitudes as high as 78° 30’ to 80° 30’, and these are: Juncus biglumis, J. triglumis, J. castaneus, Luzula Wahlenbergit, L. hyperborea, and L. nivalis. But we know also from Hart’s Report on the Botany of the British Polar-Expedition of 1875-76 (l.c. p. 19), that Juncus biglumis was collected as far north as 82° 27’, Luzula arcuata at 81° 49’, and L. parviflora at 72° 20’. And from the north coast of Siberia (Sibir. Nordkust. Fanerog. Flora) Kjellman has reported: Juncus biglumis, Luzula parviflora, L. Wahlenbergti, L. arctica, and L. arcuata. A similar high latitude is in Greenland reached by Juncus biglumis, 76° 7’; by J. triglumis, J. castaneus, and J. arcticus 70°; and Luzula spicata is reported from 76° 7’, and L. parviflora from 72° 20’. In the Alps of Switzerland Heer (l.c.) records the following species: Juncus triglumis from an altitude of 8,500 feet, J. arcticus from 8,020 feet, Luzula parviflora from 9,554 feet, and L. spicata from 9,600 feet. On Long’s peak in Colorado I collected Juncus biglumis, J. triglumis, and Luzula spicata at an elevation of 12,000 feet, while I did not observe L. parviflora above the timber- line, and never above about 10,300 feet. Finally may be mentioned that four of the species collected by the expedi- tion have been recorded from the Altai mountains, and of these Juncus triglumis and Luzula spicata have even reached the Himalayas, where they have been observed at elevations ranging from twelve to fifteen thousand feet. The prevalent arctic-alpine distribution of these Juncaceae may thus indicate either an arctic or an alpine centre. However, the species appear to thrive best in the far north, on the tundra and seashore (Juncus), in the moun- tains (Luzula). And it is a notable fact that such types of Juncus and Luzula as are alpine, but not arctic, are of a habit quite distinct from their aretic con- geners. This may be readily seen when we consider, for instance, the alpine species of Juncus: J. xiphioides, J. Drummondit, and J. Parryi from the Rocky mountains. And with reference to Luzula, the alpine element in Switzerland contains such characteristic types as L. albida,, L. nivea, L. Forstert, L. pilosa, ete. In other words, the two sections Pterodes and Anthelaia of Luzula represent the alpine element, while Gymnodes is mainly arctic. And I believe such distinction in habit in connection with a circumpolar distribution might indicate that these species belong to the arctic flora rather than being of southern, alpine origin. According to Bentham and Hooker, the genus Tofieldia contains 14 species, two of which are natives of the Andes, while the others are widely distributed in the boreal and temperate zones of the northern hemisphere. ; Three of the species are of interest to us from a geographical point of view: T. palustris Huds., T. coccinea Richards., and T. calyculata Wahlenb. Having been collected by the expedition, 7. palustris is enumerated on Table 1, and the geographical distribution covers part of the polar regions, except Nova Zembla, and we might say also Siberia, since Ledebour credits the species only to Ural (Ruprecht) and Obdorsk (Sujef). Farther south it occurs in the Alps and Pyre- nees, but is absent from Caucasus, the Altai mountains and the Himalayas. On our continent the species is distributed from Labrador to the Aleutian islands, and south to Minnesota. T. coccinea is the only species of the genus which Kjellman has credited to the north coast of Siberia; on this continent, John Macoun (Catalogue, 1.c.) gives the distribution as follows: Barren ground from lat. 64° to the shores of 92 B Canadian Arctic Expedition, 1913-18 the Arctic sea; about Jasper’s Lake, in the Rocky Mountains, and, on the west coast, north of lat. 58°, to Ounalashka, Kotzebue Sound, Chamisso Island, and Cape Lisburne, north-east coast of America. It has also been found in Green- land, on the east coast at 72° 30’ (Dusén), on the west coast at 70° (Hartz), but it is absent from Spitzbergen, Scandinavia, and Russia. In Asia it occurs, as stated above, on the north coast: the mouth of the river Lena, and of the river Olenek, besides in eastern Siberia and Kamtchatka. Toward the south it extends to the Baikal mountains. Finally, with respect to 7’. calyculata Wahlenb., this species is more southern, and has been reported from South Russia, the Alps and the Pyrenees, Ural and Kamtchatka; however, it occurs also in Sweden, but only on the island of Gotland. Tofieldia palustris and T. coccinea thus inhabit the arctic region of this continent, including Greenland. But while the former extends eastward through arctic Europe to Ural, the latter.extends westward through Siberia to the river Olenek (Long. 120° E.). With regard to their southern distribution, T. palustris does not extend beyond Minnesota on this continent, while in Europe it reaches the Alps of Switzerland; 7. coccinea does not extend beyond Canada, while in Asia it extends to the Baikal mountains. While thus the southern distribution of these species corresponds well with the northern as far as concerns Europe and Asia, their absence from the central Rocky mountains seems strange. Furthermore, with regard to the third species, T. calyculata, its occurrence in ° South Europe and Kamtchatka is difficult to combine. We have thus in the genus Tofieldia two decidedly northern species, both represented in Greenland and on this continent, and a southern, 7’. calyculata, which in the Alps of Switzerland is, to some extent, accompanied by the northern T. palustris. No doubt the present distribution points toward the north as being the geographical centre of T. palustris and T. coccinea, but it seems impos- sible to decide whether the centre was located on the American continent or in Greenland. Lloydia serotina was collected on the arctic coast; it occurs also in arctic Russia, Nova Zembla, and arctic Siberia, from Jalmal, Long. 70° 30’ E., to Pitlekaj, Long. 173° 24’ W. Farther south it is widely distributed in the Alps and Pyrenees, Caucasus, Altai mountains, the Himalayas, besides in our Rocky mountains, as far south as Colorado. Being absent from the arctic American archipelago, from Greenland and Scandinavia, Lloydia cannot be considered a circumpolar species, and the very extensive distribution in the mountains farther south may indicate a southern location of its centre, however, not on this continent. The occurrence of the plant on both coasts of Bering strait seems to indicate the road, followed from Siberia to the north of America, and from there to the Rocky mountains. The location of the original centre is difficult to decide, and it seems plausible to suppose that the species developed from two centres, one in the European Alps, another in the Himalayas. Of the genus Salix eleven species were collected on the arctic coast by the expedition, but to these I have added S. alaxensis, which was found at King point by the Gjéa expedition; furthermore, I have added S. polaris Wahlenb., according to Simmons (l.c. p. 73), who states that a specimen collected by Miert- sching at Cape Bathurst represents this species. i Three of these willows are circumpolar, viz.: Salix glauca, S. polaris, and S. reticulata; the absence from arctic Russia makes S. arctica non-circumpolar; with regard to 8. ovalifolia and S. rotundifolia, these species are not only absent from Russia, but also from Greenland and the arctic American archipelago. S. Richardsonii and S. alaxensis have been recorded from the arctic archi- pelago, besides from the American coast of Bering strait; but S. phlebophylla, S. pulchra, and S. niphoclada are known only from the arctic coast, explored by the expedition, and from the American coast of Bering strait. Finally, S. Arcive Plants: Geographical Distribution 93 B JSullertonensis, described from specimens collected at Fullerton, in the Hudson bay region, was also found on the arctic coast by the expedition; the specific name is thus one of the numerous instances that prove geographic names to be misleading, and such names ought always to be avoided. As to the southern distribution, S. reticulata has found its way to the Alps, the Pyrenees, the Rocky mountains (Colorado); it occurs also in the Altai mountains, from where S. glauca and S. arctica have also been reported. _ While thus the representation of the genus Salix, as it occurs in the arctic region, 1s well exemplified by the collection brought home by the expedition, it seems very strange that the common arctic S. herbacea L. is not among the species. This species is, as a matter of fact, reported from several stations in the’ American archipelago, and it is very common in Greenland, between Lat. 60° and 76° 7’ on the west coast, and at about 72° 25’ on the east coast (Dusén) ; furthermore, it is known from Iceland, Spitzbergen, Scandinavia, arctic Russia, Siberia (Obdorsk, collected by Sujef according to Pallas), but it is absent from the coasts of Bering strait. Further south S. herbacea occurs in Labrador, extending from there to summits of the higher Rocky mountains, from Canmore, Alta., to the Selkirk mountains (Lat. 51°) in British Columbia; in Maine it has been found in the alpine region of Mount Katahdin, and in New Hampshire on Mount Washington. In South Europe it occurs in the Alps and Pyrenees, and in Asia it has reached the Altai mountains. The distribution of S. herbacea thus agrees so well with that of S. reticulata that it was really to be expected that the species would also be found on the arctic coast. Considering, however, the species of Salix, represented on the coast, and comparing their general distribution throughout the northern hemisphere, it appears as if the genus has a most important centre of distribution in North America, and especially in the arctic region. Undoubtedly the circumpolar as well as several of the other arctic species originated in these regions, and as we know from the writings of Nathorst,! 8. polaris, S. reticulata, and S. herbacea once inhabited the lowlands of Central Europe, associated with Betula nana, Diapensia, Andromeda, Ledum, Oxyria, and several other genuine arctic plants. In other words, they are members of the glacial flora, and some are still in exist- ence on the southern mountains. One species of Populus was collected along the Hulahula river, Alaska, and along the lower Coppermine river, Northwest Territories; the specimens consisted of some dead leaves, and some wood and bark. Although the outline of the leaves does not correspond with that of typical P. tremuloides Michx., I have nevertheless ‘‘ad interim’ referred the material to this species on account of the structure of the petiole, which corresponds with that of P. tremuloides, it being very distinctly flattened. Owing to the possibility of the specimens being P. balsamifera L., as the latter species is known to extend nearly as far north, the following points relative to the distribution of both species may be given. According to John Macoun (Catalogue, lc. p. 456), Populus tremu- loides Michx. is one of the most widely distributed of the deciduous trees. It is found on all dry slopes from Newfoundland and Labrador throughout the northern forest region to Alaska. It constitutes the “bluffs” and copse wood scattered over the prairie region, and is a fine large tree in the southern part of the sub-arctic forest. Southwards it extends through the northern States to the mountains of Pennsylvania, northeast Missouri, southern Nebraska, and through all the mountain regions of the west, often ascending to an elevation of 10,000 feet above the level of the sea, to the Sierras of central California, northern Arizona and New Mexico, the high mountain ranges of Chihuahua, and San Pedro MArtir mountain in lower California. (Sargent: Silva. Vol. 9, Boston, 1896) 1 Nathorst, A. G. Uber den gegenwartigen Standpunkt unserer Kenntniss von dem Vorkommen fossiler Glacialpflanzen. (Bih. K. Sv. Vet. Akad. Hdlgr. Vol. 17, No. 5. Stockholm, 1892.) 94 B Canadian Arctic Expedition, 1913-18 Populus balsamifera L., the Balsam poplar, extends even farther north than the Aspen, P. tremuloides, and as stated by John Macoun (1.c.p. 456), it attains a great size and height as far north as the arctic circle on the Mackenzie river. From the valley of the Mackenzie river, and from the Alaskan coast, it is dis- tributed through the Hudson bay region and Newfoundland, southwest to northern New England and New York, central Michigan and Minnesota, the Black Hills of Dakota, northwestern Nebraska, northern Montana, Idaho, and Oregon and Nevada, according to Sargent.!. As far north as Fort Simpson on the Mackenzie, Lat. 62° N., Mr. E. A. Preble (l.c.p. 520) writes that the species is a stately tree, but below that point it rapidly decreases in size, and on the lower Mackenzie and Peel rivers it occurs as a small tree. “West of the Mackenzie river delta there are said to be poplars on the Turner river (Kongakut) west of Demarcation point; on the Aichillik river, south of Icy reef, on Jago river, south of Martin point; and on the Hulahula, southwest of Barter island. There are said to be no poplars on the north side of the mountains west of the Hulahula river where specimens were taken. All the streams have shrub willows a short distance back from the coast, except the Okpilak, whose name signifies “without willow (okpik)’. The Sagavanirktok has few willows, while the Canning is said to have willows 15 to 20 feet high and as thick as an ordinary stovepipe, in some of its sheltered valleys.”’ (R. M. Anderson, from field notes.) While Betula glandulosa Michx. was collected at several stations, no speci- mens of B. nana L. were discovered. And the material of the former, though very small-leaved, showéd the structure characteristic of the species, especially the branches being densely covered with resinous, wart-like glands. It seems very strange that B. nana appears to be absent from the north coast, since the two species actually cover the same area, a little farther south, from New- foundland and Labrador westward to the Mackenzie river. Fig. 4.. Thicket of small poplars, Populus tremuloides Michx., bordered by a common arcti ill Saliz arctica. Foothills of Endicott mountains, on Hulahula river, Alaska. One of the very foe straggling bunches of poplar on the arctic slope west of the Mackenzie. April 2, 1914. Dead leaves” pe still "bai on the willows, the nook being well sheltered from winds. (Photo by R. M. nderson. ] 1 The Silva of North America. Vol. 9, p. 167. Boston, 1896. Arctic Plants: Geographical Distribution 95 B Considering the geographical distribution of B. glandulosa, it extends from the north coast to both coasts of Bering strait (Kjellman), and to the northeast coast of Siberia, from where Kjellman has recorded it: Pitlekaj Long. 173° 24’ W.; according to this author the variety sibirica Ledeb. of B. nana is identical with B. glandulosa Michx. Towards east B. glandulosa is known from south Greenland, where the variety rotundifolia Regel is not uncommon, but it is totally absent from Europe. On our continent the species follows the Rocky mountains as far south as to Colorado, where it is mostly sub-alpine, and where it occurs only as the typical plant. It would thus appear as if the species devel- oped on this continent, and presumably in the northern, though not arctic, zone. Finally may be mentioned that Betula papyrifera Marsh.,the Canoe birch, was observed by Dr. R. M. Anderson ! in the Mackenzie delta; the birch was growing on the east bank of the delta, not farther than 50 miles south of Richard island. It-was not very numerous, nor very tall, not over eight or ten feet in height. No specimens of Alnus were collected, but Dr. R. M. Anderson has informed me that Alnus alnobetula (Ehrh.) Koch was found many miles north of the limit of spruce in the Mackenzie delta; according to Mr. E. A. Preble? this alder occurs throughout the region from the Saskatchewan northward to the tree limit. I presume this Alnus is the same as “‘A. ovata (Schr.) var. repens Wormskj.”’, which by Kjellman is reported from the Siberian coast, from Bering strait to Ural, and which Lange has credited to Greenland between Lat. 61° 10’ and 67° N.; with respect to its habit in South Greenland Lange writes: “Floret post foliationem, quae mense Junio incipit. Truncus ad 9 ped. altus, et fere brachii crassitie, observantibus Schiddte et Kornerup, occurrit.”’ Preble (loc. cit. p. 525) also gives the hoary alder, Alnus incana (L.) Willd., as occurring along the Mackenzie north to Peel river. According to Gray’s New Manual, the specific name alnobetula has been changed to crispa (Ait.) Pursh, since A. viridis of the Manual ed. 6 is not DC., and A. alnobetula Am. auth. is not Koch. ; “A species of alder was also found growing abundantly on the south side of the Endicott range in Alaska, along the banks of streams tributary to the east fork of the Chandalar and the Sheenjek. The inner bark of this alder is much used by the inland Alaskan Eskimos for staining the flesh side of dressed wolver- ine and occasionally caribou skins. The outer layer of bark is scraped off, the inner bark removed, cut into small bits, moistened with water and applied to the skins, the resulting colour being a burnt sienna similar to the ochre-stained skins of the Siberian Eskimos.’”—(R. M. Anderson, notes). ; Covering the immense area in the arctic and southern zones, as given in the accompanying table (Table 1), we have in Polygonum viviparum and Ozxyria two genuine arctic types of circumpolar distribution. On the north coast of Siberia they both were collected by Kjellman, Oxyria between Long. 68° E. and Long. 173° 24’ W., and the Polygonum between Long. 80° 58’ E. and Long. 173° 24’ W.; Oxyria was even among the plants which this author collected on the northernmost point of Asia: Cape Tscheljuskin, Lat. 77° 36’. They are also among the plants growing nearest the Pole, both having been recorded from Lat. 82° 27’ N. (Hart); in the mountains further south they ascend to an eleva- tion of from 9,500 to 10,000 feet in the Alps of Switzerland (Heer) ; in the Rocky mountains, Colorado, they both occur at 14,000 feet, and in the Himalayas they even reach an elevation of from 10,000 to 15,000 feet (Hooker). Oxyria is monotypic; P. viviparum is of the small section Bistorta, of which, furthermore, P. Bistorta was collected by the expedition; a near ally of the latter is P. bistortoides Pursh, a native of the Rocky mountains, and frequent in the alpine region as far south as Colorado. But there is another species of this 1 af; ,V. My Life with the Eskimo. New York, 1813, p. 444. 2 rene ‘eal Tewancigation of the Athabaska-Mackenzie Region. (North Am. Fauna. No. 27, U.S. Dept. of Agric., Washington, 1908, p. 525.) 96 B Canadian Arctic Expedition, 1913-18 section which is much more closely related to P. viviparum, and which shows the same peculiar transformation of the flowers into bulblets, P. Macounti Small. This interesting species was detected by James M. Macoun on St. Paul island, Bering sea. Oxyria and Polygonum viviparum evidently originated in the polar regions; during the glacial epoch they became distributed farther south, where they are still in existence, especially in the alpine regions. P. Bistorta, on the other hand, may be regarded as a southern type, as one of those which accompanied the arctic flora on its retreat to the north. The fact that P. viviparum and its near ally P. Macounti, besides P. Bistorta, and P. bistortoides, are all indigenous to this continent seems to indicate that the section Bistorta must have an important centre of distribution in the northernmost part of America, especially in the northwestern corner. Koenigia islandica L. was not collected by the expedition, and in North America it is known only from some of the islands of the arctic archipelago,- from Unalaska, and, according to Macoun, from ‘‘Moist mountains north of Smoky River, Lat. 54° (Drummond).”’ It is one of the very few arctic plants which are annual, and its geographical distribution may be of interest to students of arctic botany. The species is almost circumpolar, since it has been recorded from Greenland, Spitzbergen, arctic Scandinavia, arctic Russia, Nova Zembla, and, as mentioned above, from some of the islands of the arctic archipelago. But it is not among the plants which Kjellman has recorded from the north coast of Siberia, and Ledebour cites no stations from arctic Siberia either; farther south Koenigia islandica is known from Altai and Baikal. On the east coast of Greenland it reaches as far north as Lat. 74° 30’ (Dusén), and it is not uncommon on the west coast between Lat. 60° and 72° 48’. Koenigia is not monotypic, a second species being indigenous, or let us say, endemic to the Himalayas, and this second species is also an annual. We have thus in the genus Koenigia two species, exhibiting such extremely different distribution as almost circumpolar ( K. islandica), and endemic to the Himalayas (K. nepalensis Don). The Caryophyllaceae are well represented in the arctic region, and of the 13 species known from the north coast, six are circumpolar; according to Hart (l.c. p. 19), some of these plants are among those reported from the farthest north, for instance, Cerastiwm alpinum Lat. 82° 50’; Alsine verna and Stellaria longipes Lat. 82° 27’; Lychnis apetala Lat. 81° 52’; L. affinis Lat. 81° 50’; Silene acaulis Lat. 81° 40’; Stellaria humifusa Lat. 78° 18’; Alsine arctica Lat. 72° 20’, etc. In the Alps of Switzerland Silene acaulis ascends to an altitude of 9,500 feet, and Cerastium alpinum to 8,500 feet, according to Heer (1.c.); in the Rocky mountains of Colorado I collected these two plants at an elevation of between 12,500 and 13,000 feet. While none of these arctic species have been found in the Himalayas, Hooker (1.c.) mentions that the high-northern Cerastium trigynum does occur in West Himalaya at an elevation ranging from 11,000 to 17,000 feet. With the exception of the monotypic Merckia, the other genera are widely distributed farther south, and while these: Silene, Lychnis, Stellaria, Cerastium, and Alsine, are mostly lowland plants in the Old World, Europe and Asia especially, they are in North America represented by species which are princi- pally mountain plants. For instance, of the 42 species of Silene credited to North America according to the Synoptical Flora, 12 species inhabit the Coast range (California), 6 the Cascade mountains, 10 the Rocky mountains, and 4 the Appalachian mountains. Of the 12 species of Lychnis, 5 species inhabit the Rocky mountains, and one the Appalachian mountains; of Cerastium, 17 species are known from this continent, and 5 of these inhabit the Rocky mountains, and 2 the Appalachian mountains; furthermore, with regard to Stellaria, 23 species are recorded from Arctic Plants: Geographical Distribution 97 B North America, of which 3 inhabit the Coast range, 1 the Cascade mountains, 6 the Rocky mountains,.and 8 the Appalachian mountains; finally, of our 19 species of Alsine, 2 occur in the Coast range, 3 in the Cascade mountains, 5 in the Rocky mountains, and 4 in the Appalachian mountains. The general distribution of the number of species may be seen from the accompanying table (Table 3). Considering the fact that the column ‘‘Moun- tains of North America’’ comprises the Coast range, the Cascade mountains, the Rocky mountains, and the Appalachian mountains, the number of species is remarkably small as compared with Altai, and especially with Caucasus, so far as concerns Silene and Cerastium especially. The number of species from the Alps and Pyrenees is also small as compared with Altai and Caucasus. With regard to the number of species endemic to Caucasus, there are 15 species of Silene, 9 of Cerastium,'7 of Alsine, and 2 of Stellaria; endemic to Altai are 4 species of Stellaria, 2 of Cerastium, and 1 of Lychnis. a ‘ z a ) 8 > TABLE 3. Zz 2] a g 3 | 3 3 | & | Distribution of some Caryophyllaceae in the | ° a oI aS g g B mountainous and arctic regions of a aS) 3 7 aA ss 8 o 2 both Worlds. ae | § a a 3 a a ea| es 8 g 8 8 8 2 sa/ 2) eg |e) 6) ete le = < 1S) < Ay < Wissdectendd Pie oes 3 3 1 AT SIRE Li cicion dsciere a Reet ks @ ABIANS oles oes w seers aH 14 4 12 8 6 3 2 4 SEGUE GR Ua assess see-c\ ecu tea evade. aes cayace donee Mesias Veaes Bete eE 18 14 6 bia eee 8 6 2 CPP ce dni Sees Camarasa Ra ieee eta esa teens 7 9 26 4 4 5 4 2 And as may be seen from the same table (Table 3), the contingent of these mountains to the arctic Flora is very small; still, as has been mentioned above, the Caryophyllaceae are quite frequently met with in the arctic region, and six of those collected by the expedition are even circumpolar and extend very far north. It would thus appear as if these circumpolar species besides Stellaria longipes originated in the arctic regions, where they are better represented than anywhere else farther south. However, Halianthus, although being circumpolar, must not be included, since it is a lowland plant inhabiting the seashore farther south; the migration of this plant to the polar regions must undoubtedly have taken place at the same time as so many of the other southern species accom- panied the arctic on their retreat to the north. Merckia physodes is undoubtedly a genuine arctic type, known only, so far, from Alaska, Eschscholtz bay and Norton bay to Point Barrow, Mackenzie delta, and northeastern Siberia, Konyam bay, and mouth of the river Kolyma. Alsine arctica and A. macrocarpa are both recorded from the north coast of Siberia by Ledebour and Kjellman, and the former has also been collected in the Altai mountains and in Kamtchatka. ; ; Very remarkable is the distribution of Cerastium maximum L. (C. grande Greene). This species has been reported from arctic Siberia: Mouth of the river Kolyma (Angustinowicz), and Pitlekaj (Kjellman); furthermore, from several stations in the Altai and Baikal mountains, from the arctic coast of this continent, from Yukon (Hunker creek), and finally from as far south as Caucasus. 24657—7 98 B Canadian Arctic Expedition, 1913-18 By considering the table (Table 3), it appears as if Caucasus and the Altai mountains constitute a most important centre for the development and dis- tribution of the Caryophyllaceae, especially the genera Silene, Alsine, and Cerastitum; however, only so far as concerns the mountainous species, for the majority of the members of the family are lowland species indigenous to the temperate zones of Europe and Asia. \ Among the Ranunculaceae, Caltha palustris is in the arctic regions nearly always represented by the variety radicans (Fr.) Hn., and is cireumpolar, being absent only from Greenland and Spitzbergen; it is not by Ledebour credited to arctic Siberia, but Kjellman collected the plant at Irkajpi, Long. 179° 25’ W., on Preobrascheni island, Long. 113° 10’ E., and at Dickson’s harbour, Long. 80° 58’ E. Being by far more abundant and typically developed in the temperate regions of both Worlds, and especially in the lowlands, the geographical centre must be sought in these regions, in Europe or Asia. TaBLe 4. Atlantic | Rocky Facific Geographical distribution of the genus Anemone in North America.| slope mtns. slope Sectio: Pulsatilla— A. hirsutissima (Pursh) MacM..........0 00.006 e cece eee fer eee e es Jt ah ataghin Sectio: Preonanthus—. A. Occidentalts- Watss.cuincciuranae cunpontinasaey aren nasans ee bar RED - * Sectio: Anemonanthea— A. parviflora Michx................ Piet sie, Rta heh Sith tie AteeBan Yale ts ie - Aj, Drum Monde: Wats vescecccinacammunaaweweenmanniey Aatud wee Saye Paeranisoinghe - A CECODCIIO ATG ssscrcrintndiatcs, PR gcingmsdclges Geiyeauauioe her katea allaveevedmes any (ener ene A. sphenophylla FOR ie cucn chug ey caer eee AWE ee: Os a cleseungaads * * A caroliniang, Waltss career uk Watetesn th Lege ahidocu Beate acws aL eee eee ee Ardeltordea TIOO ess saccade hata “Wuesioaithgy Bioaaediae Geb duntavieatl oemnensesie lkavoseiucs ae * A. Richardsonii Hook.....00..0 0.0000 cc ccvcceeseveeeee cesteteeceethecccecee. . = BAU ss UUM UCL OULD so cated ctr pab sey ode saya tate ot the tenn Dah onde Sade TD Ih saanneaaigea tell ea isaseisevs AAS, ONCG O05 G Tiare secs cutscnsigeais steentem taupe aa aaah en tghi aca Paw a lage ASR epee HL ee i caeetmnt | heed oceceboand * Ata toa IR uli cs ainenae Sec sacl cas aeceiviga Roan da sadtamomsenibia cecal Ga Ase entice * Be VLOG Nines a eh 09 65-99 BR EAI AERIS BA DV RNS DAP RR dma WA ak Beart nae tine | bseencrnsehed camsiedh Sectio: Anemonospermos— SA We ha Unt oui atad naar oan gitar kad ssehalnae MeN wey * * * AgteiOnensis POT bsrsrrya arava wie sareradir ast Vi arg MPEeee ei Mt wesad bisa eameaad co} aoe rmanese * Ap cylindrica Gis, 5 coaiee ete rare Dee Nae Hei ire eae asennad Hietgahaase * Foal dig ee cae DAs Baie Mewen Ta Maa sca a8 gash a haar snc butthead Decade emcee ip ileal cen GA COMO MENGES ei acc tea saiboei anos yaad cancaors jeatiecdash bli a aus Dusebaneeaidmeateeite Rehan * : et ees Sie Sectio: Omalocarpus— : AB SHR EUS SUPE Ore Nia acest Sic ooveriy techs bccn Rs aw be a2 oa elves Segoe regener * * The genus Anemone is in the polar regions represented by the sections: Pulsatilla, Preonanthus, Anemonanthea, Anemonospermos, and Omalocarpus ; no other se¢tions are represented on this continent. However, the distribution is so extremely scattered in the arctic regions that it seems necessary briefly to discuss the distribution of the genus farther south before outlining the occurrence Arctic Plants: Geographical Distribution 99 B ‘n the north, in the arctic regions. De Candolle is the author of these sections,! and there is only one more, Pulsatilloides, with two species, both from Cape of Good Hope. From the table (Table 4), it will be seen that A. hirsutissima is the only member of the section Pulsatilla represented on this continent. According to Gray (Synopt. Flora l.c.) A. hirsutissima is identical with A. patens L. var. Wolfgangiana_Regel, which is also recorded from several stations in the Old World, viz.: Sweden, Germany, Russia, Altai, and Baikal; the only place, how- ever, where it extends beyond the arctic circle is on the north coast of this continent. The section is much better developed in Europe, from where six species have been recorded, three of these being also indigenous to Siberia. Anemone parviflora extends from Labrador and Anticosti to the Pacific slope, while A. Drummondii does not occur on the Atlantic slope. With regard to A. Richardsonit, this has been collected on the arctic coast, on the islands of the arctic American archipelago, and at a very few stations on the west coast of Greenland; by Macoun the distribution in Canada is given: “Shores of Hudson’s Bay; barren ground, Rocky Mountains, from Lat. 55°-68°, in wet, mossy ground (Richardson, Drummond); York Factory (R. Bell).”’ Moreover, it occurs on both coasts of Bering strait: Port Clarence, St. Lawrence bay and Konyam bay. (Kjellman); it has also been found in ‘‘terra Tschuktchorum” (Maydell). None of the other members of Anemonanthea extend to the arctic region of this con- tinent, while in Europe A. nemorosa L., and A. ranunculoides L. have both been recorded from arctic Scandinavia. Among the southern types of this section we notice (Table 4) that there is on this continent a small group of species which show a distinct analogy to a corresponding group in Europe and Asia. I think especially of A. quinquefolia, A. oregana, A. Lyallii and A. trifolia, and the northern A. Richardsonti, on this continent, and of A. nemorosa L., A. ranuncu- loides L., and A. trifolia L. in Europe and Asia. A. trifolia is thus common to both Worlds; on this continent it is distributed from ‘Mountains of 8. Pennsyl- vania to Virginia,’ while in Europe it occurs in Tyrol, Steyermark, Karnthen, etc.; in Siberia it is replaced by A. reflera Steph. It would thus appear as if these two groups of analogous species had developed independently on both continents, since it would be very difficult, if not impossible, to connect the stations so remote from each other. In other words, the American element may have had its origin on this continent. The section Anemonospermos is not represented in Europe, but in Siberia by a single species, A. dichotoma L. With respect to the section Omalocarpus, only A. narcissiflora L. is represented on this continent, and in Europe, while two other species are known from Asia: A. baikalensis Turcz. from Baikal, and A. sibirica L. from the region of Jenisei; the section is undoubtedly of southern origin, but may have developed from more than a single centre. With regard to the arctic representation of the genus, it would appear as if the American element did not originally develop in the polar region, but farther south, in the Rocky mountains, from where they migrated to the north when the ice receded, A. Richardsonii even entering Greenland. In Europe the two low- land species did evidently reach the arctic region at the same time, while their original home can only be defined as Central Europe. , In comparing the arctic American element of Ranunculus with that of Europe and Asia, we notice at once the presence of certain genuine American types on our coast, absent from the Eurasian; conversely some Eurasian types present on the north coast of the Old World, but absent from the North American. We have thus in arctic Scandinavia and arctic Russia four distinct types: R. auricomus L., R. acris L., R. repens L., and R. sceleratus L., none of which are indigenous to North America except perhaps the last one; moreover, R. glacialis L., which is known from the east coast of Greenland (about Lat. 74° 1 Regni vegetabilis systema naturae. Vol. 1, Paris, 1818, p. 189. 24657—73 100 B Canadian Arctic Expedition, 1913-18 30’, Dusén), but which has not been found, so far, on the North American coast. With regard to the arctic Siberian element, Kjellman (1.c.) has recorded nine species from the north coast, principally collected by himself on the Vega expedition. These species are: R. Chamissonis Schl., R. Pallasii, R. sulphureus, R. nivalis, R. pygmaeus, R. hyperboreus, R. affinis, R. acris, and R. lapponicus. Concerning the arctic American element, this consists of six species, namely: R. Purshii Richards., R. Cymbalaria Pursh, R. Sabinit R. Br., R. occidentalis Nutt., R. lapponicus L., and to some extent, R. affinis R. Br. Ranunculus occidentalis Nutt., a near ally of R. recurvatus Poir., is a native of open woods and low ground in the northern Rocky mountains, from where it extends to the Alaskan coast and islands, and south to the borders of California. The variety robustus Gray has been collected on the Alaskan islands, and on the north coast. With regard to the representation of the genus throughout the northern hemisphere, the Old World element is quite distinct from the North American. On this continent 53 species are enumerated in the Synoptical Flora, not includ- ing 6 introduced species, and not either the genera Kumlienia, Arcteranthis, Cyrtorhyncha, and Batrachium. And among these species of Ranunculus some certain groups are much better represented on this continent than in Europe and Asia. For instance, the section comprising R. ambigens Wats., which corresponds to the Eurasian R. Flammula alliance; furthermore, the sections of R. adoneus Gr., and of R. abortivus L., which do not seem to have any analogous representations in the Old World. On the other hand, the section comprising R. hispidus Michx. is well exemplified in the Old World, and more extensively so than on this continent. The circumpolar species: R. hyperboreus, R. pygmaeus, R. nivalis, and R. lapponicus, besides R. Pallasti, R. sulphureus, R. Sabinii, and AR. affinis are, as may be seen from Table 1, much more widely distributed in the polar regions than farther south; some few have extended as far south as the Altai mountains, and some even to the Himalayas. But their principal range lies within the arctic region where they undoubtedly originated. R. ‘Cymbalaria does not properly belong to the arctic region, even though it has reached the west coast of Greenland at a very few stations only; the distribution of this species is mainly “sea-shore and margins of salt-ponds in the prairie region of this continent;’”’ however, it has also been recorded from several stations in the Hudson bay region, and even farther north. While thus the arctic species, with the only exception of R. Sabinit, have also been recorded from the Old World, it seems somewhat strange that the American alpine element of the genus is not represented; and the same ‘is the case of the Eurasian alpine element, for the few species enumerated above as having found their way to the arctic region, R. auricomus, R. repens, etc., are lowland types, and the extension of their southern range northward is evidently of a relatively recent date. The circumpolar Papaver nudicaule is a genuine arctic type which during the glacial epoch became widely distributed in the higher mountains farther south; and although still in existence in the Altai mountains, Dovre mountain in Norway, and in the Rocky mountains as far south as Colorado, there ig no trace of the species in the Alps or Pyrenees. It is one of the three plants which have been found north of Lat. 83°, Saxifraga oppositifolia and Alopecurus alpinus being the two others; it may also be mentioned that Papaver nudicaule occurs in western Thibet at an elevation of 16-17,000 feet. Among the Cruciferae we meet with types of very distinct geographical distribution: I. Circumpolar: Draba alpina, D. nivalis, D. fladnizensis, and D. hirta; Eutrema, and Cardamine pratensis. II. Arctic, but not circumpolar: Lesquerella, Cochlearia, Draba corymbosa, Braya, Hesperis, Cardamine digitata, and Parrya. Arctic Plants: Geographical Distribution, 101 B III. Southern, alpine: Erysimum, Thlaspi. IV. A weed: Sisymbrium. Of these the circumpolar, and the arctic but not circumpolar, with the only exception of Cardamine pratensis, may be considered as having originated in the polar regions. Erysimum inconspicuwm belongs more properly to the Rocky mountains where it is associated with several species and more widely distributed: Cardamine pratensis is circumpolar and, moreover, it is widely distributed through the temperate regions of both Worlds. In the arctic region it has only two congeners: C’. bellidifolia L., which is circumpolar, and C. digitata Richards., which seems to be a rare plant, known only, so far, from the islands of the arctic American archipelago, from both coasts of Bering strait, from the Yukon- Alaskan boundary, Long. 141° W., Lat. 60.02 N., from the Hudson bay region and from the territory explored by the expedition. But farther south the genus 1s quite rich in species, annual or biennial, as well as perennial. Some of these extend as far south as Cape of Good Hope (C. anteniquana Burch. and C. africana L.), and Tierra del Fuego “fere sub nive” (C. glacialis DC.). And with respect to C. pratensis, this species shows a wide distribution in the moun- tains of Europe and Asia, Caucasus, the Himalayas, Altai, and Baikal moun- tains; it is, however, absent from the Rocky mountains. A peculiarity of this species is that it sometimes does not develop mature seeds, not even in Central Europe, but that reproduction may take place by means of bulblets developed upon the leaves. In the arctic region this species sometimes does not even reach the flowering stage, as pointed out by. Nathorst, for instance in Spitz- bergen, and this peculiarity it shares with Empetrum, Petasites frigida, Ranun- culus Pallasii, and R. hyperboreus. Nevertheless, these plants are quite widely distributed in Spitzbergen, and Nathorst explains this by considering the plants to represent remnants of an earlier period when the climate was sufficiently warm to enable them to develop fruit and mature seeds. However, judging from the wide geographical distribution and the abundant representation of the genus in the south, it seems most natural to consider the species as being of southern origin, but that it partook in the migration towards north when the arctic flora retreated. But it is impossible to decide with any certainty the location of the original centre of this species; it may, however, have been in Central Europe where the species is most abundant and associated with a number of allied species, and also with several of the closely allied genus Dentaria. In the United States the species is not so common and mostly restricted to the Northern Atlantic States; on the other hand, there are some allied species in the Western States, for instance C. rhomboidea DC. and C. cordifolia Gray in the Rocky mountains, besidts that several species of Dentaria are known from the Atlantic and Pacific slope. In considering the arctic element in Europe and Asia the Cruciferae are in Finmark represented by not less than 24 species, 8 being introduced weeds, however; in Russia there are 28 species, and in Siberia 20, according to the enumeration in Kjellman’s paper (Sibir. Nordkust. Fanerog. Flora). By looking over the representation of the family on this continent, it is interesting to see how several genera have developed here, being endemic to North America (Thysanocarpus, Lesquerella, Physaria, Streptanthus, Caulanthus, Thelypodium, etc.); and among the genera, which are common to both Worlds, several have in America given rise to a relatively large number of endemic species, for instance, Draba with 20 species endemic, Arabis with 30, Dentaria with 9, Hrysimum with 7, ete. But from the geographical table (Table 1) it is surprising to see how ex- tremely few alpine, I mean southern alpine, types have entered the polar regions; for the circumpolar must be left out of consideration, and with respect to the simply arctic species, none of these bear any evidence of ever having been more 102 B Canadian Arctic Expedition, 1913-18 generally distributed south of the arctic region; moreover, the number of species that occur in Greenland and on this continent only is extremely small: Lesquer- ella, Draba aurea, Hesperis, Arabis Holboellit, and A. Hookert. By comparing the geographical distribution of the genus Sazifraga, we have seen that not less than 7 out of the 15 species collected are circumpolar; and with the only exception of S. rivularis, they are, furthermore, widely distributed farther south. Engler,! who has arranged the species in very natural sections, attributed the following geographical distribution to those of the species which occur within our region, i.e., the one explored by the expedition. Nephrophyllum: Of the 19 species, representing this section, 10 occur in the mountains of Spain, south of the Pyrenees, and 6 of these are endemic to these mountains; among the others some few are known from the Alps, Caucasus, the Himalayas, eastern Siberia, the Rocky mountains, and Alaska. The section Hirculus contains 11 species, 10 of which are endemic to the Himalayas; the eleventh species, S. Hirculus, shows, as may be seen from the table, an extraordinarily wide distribution throughout the northern hemisphere. Boraphila with 23 species is mainly Siberian, about 16 species being charac- teristic of the districts between Altai and Kamtchatka, besides western North America; some other species are more widely distributed on this continent, and 9 of these are endemic; several species are also known from the Himalayas, among which 3 endemic. The section is, furthermore, widely distributed in the arctic region. While 36 species are attributed to the section Dactyloides, most of these are from Spain, Atlas, and the Canary islands; some few are known from the Car- pathian mountains and the South American Cordilleras; a very few species occur in the Asiatic mountains, but none in the Himalayas. With regard to the section Trachyphyllum, most of the species are Himalayan, and only a very few have been reported from Altai, Kamtchatka, western North America, and Central Europe. Of the four species representing the section Porphyrion, three are confined to the mountains of Central and South Europe, while the fourth one, S. oppositi- folia, shows an enormous distribution through the northern hemisphere and is, besides, circumpolar. The data regarding the distribution of the sections, of which species were collected by the expedition, thus indicate the difficulty in locating the actual centres of their distribution, whether the species originated in the north, notably the arctic region, or in the south. However, with reference to S. radiata Small (S. exilis Steph.), the very limited distribution points towards the centre having been Iqcated in northeastern Siberia, where it is accompanied by a near ally, S. stbirica. S. refleca Hook., being exclusively arctic and confined to the coast between the Mackenzie and Coppermine rivers, must have originated there. S. aestivalis grows on the north coast of Siberia, but is seemingly rare there; from there it extends to Altai and Baikal, and eastward to Kamtchatka, the coast of Man- churia, and St. Lawrence island. On this continent the species has been reported from Alaska, together with the very local S. Nelsoniana, as well as from the Rocky mountains, as far south as Colorado, and the Cascade mountains. Being apparently very rare in the arctic region as compared with its distribution farther south, and especially in eastern Asia, I presume the centre of both.S. aestivalis and S. Nelsoniana must have been in northeastern Asia. With respect to S. bronchialis, the very wide distribution of this species in the arctic region of Russia and Siberia, and being so very rare on this continent, absent from Greenland and the arctic American archipelago, seems to indicate that its geographical centre may have been located in Siberia, but not necessarily within the arctic region since the species seems to be quite well distributed 1 Monozraphie der Gattung Saxifraga. Breslau, 1872. Arctic Plants: Geographical Distribution 103 8B farther south, for instance, in the Altai and Baikal mountains, Davuria, and Kamtchatka. On this continent it does occur on the arctic coast, but only at a very few stations; from there it extends south to the Rocky mountains in Colorado. Although a near ally of S. tricuspidata, the geographical distribution of these does not coincide, for the latter is common in Greenland and on the islands of the arctic American archipelago; it has also been recorded from the Appalachian mountains and the Hudson bay region but not from any stations in the Old World. We have thus two closely allied species, one of decidedly Siberian origin and another, confined to this continent, but both occurring in the arctic region and, so to speak, meeting each other on the coasts of Bering strait. The present distribution of S. tricuspidata, notably in Greenland, where it abounds between 60° and 78°18’ N.L. on the west coast, and from 60° to 76° N.L. on the east coast, besides that it seems to be quite frequent on the arctic islands, makes the suggestion very probable, that the centre may be sought there, in the north eastern arctic corner of this continent. A very strange distribution is exhibited by S. aizoides, viz.: many stations in Europe, from the arctic region (Spitzbergen and Scandinavia) south to the Pyrenees and the Alps of Italy and Switzerland; in America from the arctic coast south to 51° N. L. in the Rocky and Selkirk ranges, but absent from Asia except a very few stations in Nova Zembla and Ural. As far as concerns the arctic area covered by this species in Europe compared with the corresponding latitude on this continent, including Greenland, the American area is consider- ably wider. But the much larger area occupied by the species in temperate Europe, besides the more frequent occurrence of the species, does indicate, I believe, that the species originated there, and possibly in the Alps or Pyrenees. Being almost circumpolar, the only districts uncovered being arctic Scan- dinavia and the Russian coast, Sazifraga flagellaris may nevertheless be con- sidered as a truly arctic type which originated in the polar regions. It is widely distributed on the arctic coast of this continent, including Greenland; it is known from Spitzbergen and Nova Zembla, and from several stations on the Siberian coast. Farther south it is known from the Rocky mountains (Colorado), Caucasus, Altai and Baikal mountains, and the Himalayas. However, it is true that most of the species of the section Trachyphyllum are Himalayan, as pointed out by Engler (1.c.), and, moreover, this peculiar little species is in the Himalayas associated with two others, S. Brunoniana Wall., and S. pilifera Hook. fil. et Thoms., which exhibit exactly the same habit as S. flagellaris, viz.: filiform stolons above ground being developed from the axils of the rosette- leaves, and terminated by a gmall, leafy rosette. But it would be difficult to imagine that these mountains, the Himalayas, should be the actual centre of a distribution so enormously wide in the polar regions. I am more inclined to believe that S. flagellaris, like most of the circumpolar species, originated in the arctic regions, and that it reached the Himalayas by way of the Altai and Baikal mountains. The association with the two analogous types may simply indicate a recent geographical centre of this interesting little group of species, extending as far west as Caucasus. The remaining species of Sazifraga: S. rivularis, S. cernua, S. Hirculus, S. nivalis, S. hieraciifolia, S. decipiens, and S. oppositifolia are circumpolar, and of these S. Hirculus is the only one which has no immediate allies outside the Himalayas, and the species itself is also widely distributed in these mountains, ranging from western Tibet (4,500-5,000 m. above sea-level) to eastern Hima- layas, at the same altitude. Considering the fact that the section Hirculus, containing 11 species, has 10 of them endemic to these mountains, it is natural to suppose that the Himalayas constitute an important, or perhaps the only centre, where the section has developed further. But the actual centre of dis- tribution of the species S. Hirculus may not have been located there; it seems more natural to suppose that similar to most of the other circumpolar species d 104 B Canadian Arctic Expedition, 1913-18 S. Hirculus originated in the arctic regions. For nowhere where the species occurs does it exhibit any pronounced tendency to vary, except with regard to the foliage: “oblongo-lanceolata,” “linearia,” “spathulata,” etc. By some authors, Lange, for instance (Consp. Fl. Groenl.), the arctic plant is considered as representing a variety “alpina Engl.” from the Himalayas, but the deviation from the type depends merely upon a dwarfed, more condensed growth, the leaves a little broader, etc. The arctic plant resembles the alpine; it is able to produce flowers and ripen the seed even as far north as Spitzbergen, according to Hesselman (l.c.), although probably not every year. With respect to S. rivularis and S. cernua, both being also widely distributed farther south, especially the latter, these species are well adapted to the arctic climate, by developing bulblets which, in the latter, have almost entirely re- placed the flowers; in this respect S. cernua resembles the arctic form of 8. stellaris, as described. The geographical centre has undoubtedly been within the arctic regions. The prevalence in the north, together with the circum-. polar distribution, seems to indicate that S. nivalis and S. hieracitfolia are also arctic types and that they originated in the polar regions. But with regard to S. decipiens, it seems somewhat doubtful whether an arctic centre could be credited to this species. We should remember that the section Dactyloides is decidedly southern, and S. decipiens is, so far as we know, about the only one which occurs in the polar regions and is circumpolar. It has, however, two near allies, the distribution of which reaches far north, namely, S. hypnoides L., which is common in Iceland but more widely distributed in southwestern Europe, and S. silenaeflora Sternb., known only from arctic and sub- arctic North America, viz.: Hudson bay region and Alaska. As far as concerns S. decipiens, this polymorphic species is best developed in the mountains of Central Europe, but is totally absent from the Asiatic mountains. Engler (1. c.p. 186) has enumerated several varieties, among which “oroenlandica”’ is credited to the arctic regions, and also to Norway, Iceland, and Great Britain; two other varieties, “caespitosa” and “uniflora,’’ are also credited to the northern regions. But the more evolute forms “vulgaris” and “quinque- fida” are exclusively southern. It would thus appear as if the species had developed in the south, Central Europe for instance, where it is best represented and best developed. Wherever it is met with in the polar regions it is always of a stunted growth and few-flowered, which might indicate that it is a foreigner in these regions, but having adapted itself remarkably to the change of condi- tions, and having acquired such wide distribution so as to become circumpolar. Otherwise with regard to 8. oppositifolia. This species being circumpolar and so well adapted to the arctic climate may, ir spite of the enormously wide distribution farther south, be considered as a truly arctic type. The other members of the section Porphyrion occur only in the highest mountains of Central Europe, but one of these, S. biflora All., has also become distributed as far north as arctic Russia. From a geographical point of view the genus Dryas is very interesting, and in some respects quite remarkable. D. octopetala and D. integrifolia were both collected by the expedition, while the more southern species D. Drummondii Hook. was not met with. As may be seen from Table 1, D. octopetala is not only circumpolar but is also widely distributed farther south, notably in Iceland, Scandinavia, the Alps and Pyrenees, Caucasus, Altai and Baikal mountains, the Rocky mountains south to Colorado, and the Asiatic coast of Bering strait, viz.: St. Lawrence and Konyam bay. D. integrifolia, on the other hand, is in the arctic region confined to the shore explored by the expedition, the American archipelago, Greenland, and the coasts of Bering strait. In Canada D. integrifolia has been recorded from Island of Anticosti, Labrador, the coasts of Hudson bay, extending westward to Bering strait; its southern limit in the Rocky mountains is: Summit of Moose mountain, Elbow river, at an elevation of 7,500 feet. Arctic Plants: Geographical Distribution 105 B But of these D. octopetala is very rare on the islands of the American archi- pelago (Boothia Felix), and on the west coast of Greenland it has only been found at a very few stations hetween 76° and 79° N. L., while it is frequent on the east coast between 70° and 76°N. L. D.integrifolia, on the other hand, is very .common on the west coast of Greenland, between 60° and 76° N. L., while it has only been found at a few stations on the east coast, between about 70° and 74° N. L. (Dusén and Hartz); from the islands of the archipelago it has been recorded from many stations. On the shore between Point Barrow and Bathurst inlet D. octopetala was found only near Sadlerochit river, Alaska; by the Gjéa expedition it was found at King point, together with D. integrifolia and the forma intermedia Nathorst. D. integrifolia was, on the other hand, collected at many stations along the entire coast explored by the expedition; the intermediate form was found at Port Epworth harbour. Farther westward, at Port Clarence, D. integrifolia was observed by Kjell- man to be quite frequent, while the other species seemed to be absent; but on the Asiatic coast of Bering strait Kjellman found both species, and D. octopetala was very abundant, the other species very rare. Finally, on the north Siberian coast from 173° 24’ W. L. to 68° E. L. D. octopetala is the only species, so far observed, according to Kjellman (Sibir. Nord-Kust. Flora L.c.). With regard to the third species, D. Drummondit, the distribution of this is on this continent confined to the Canadian. provinces. In his Catalogue of Canadian Plants (1.c.), John Macoun states that it never grows on mountain slopes, and that it is distributed from Quebec westward to British Columbia, and extends as far north as the shores of the Arctic sea, according to Richardson. Moreover, it occurs in eastern Siberia, along the Aldan river (Turczaninow). With regard to the Siberian plant Ledebour (FI. Ross. lc. Vol. 2, p. 21) makes the following statement: ‘‘Specimina fructifera, quae benevolentiae cl. Turcz. debeo, ab americanis a cl. Hooker mecum communicatis non differunt nisi laciniis calycinis paulo angustioribus et longioribus.” We have thus in the genus Dryas three species, the only ones of the genus, viz.: D. octopetala which is circumpolar-alpine, D. integrifolia which is arctic- alpine, and finally D. Drummondii which is only exceptionally arctic, and which furthermore does not seem to be alpine farther south. The centre of distribution of D. octopetala and D. integrifolia may have been within the polar regions, from where they migrated towards the south during the glacial epoch. Of these the former became thus widely distributed in the mountains of both Worlds, except the Himalayas, while D. integrifolia did not extend beyond the mountains of this continent. D. Drummondit, on the other hand, evidently had its centre south of the arctic regions, in the Canadian Rockies evidently, and may have developed much later than the others. D. octopetala and D. integrifolia, especially the former, are still, as we know, repre- sented on the higher mountains as remnants of a glacial flora. The limited occur- rence in the arctic region shown by D. integrifolia well corresponds with its present distribution farther south on the American continent, but, as mentioned above, this species and D. Drummondii have extended their range in a western direction to the eastern part of northern Asia. But with regard to D. octopetala, - the wide range which it occupies throughout the mountainous districts of both Worlds is seemingly in good accordance with a former, equally extensive dis- tribution in the polar regions. _ Another interesting distribution may be illustrated by the genus Sieversia, and although only one species, S. glacialis, was found by the expedition, it may be appropriate to give a general view of the distribution of the genus “‘in toto.” The following species occur on this continent: S. glacialis R. Br., S. triflora R. Br., 8S. Rossii R. Br., S. calthifolia Menz., and S. radiata (Michx.) Greene. Of these, S. triflora is not arctic, but extends from Labrador and Newfoundland to British Columbia, and follows the Rocky mountains south to Colorado 106 B Canadian Arctic Expedition, 1918-18 S. radiata grows in the Alleghany mountains, North Carolina, and a variety Peckiit Gray has been recorded from alpine tops of the White mountains, New Hampshire; S. Rossii is common in the alpine region of the Rockies in Colorado, extending northward to the arctic shore and the archipelago, and westward to Alaska; S. glacialis is known only from the arctic seashore west of the Mackenzie river; S. calthifolia is known from the northwest coast, Unalaska and Sitka. According to Ledebour, S. glacialis and S. Rossii occur in eastern Siberia, S. calthifolia in Kamtchatka. Beside these a fourth species is enumerated by Ledebour (Flora Ross. l.c.), S. anemonoides R. Br., also from Kamtchatka. Finally, there are two species in the Alps and Pyrenees, S. montana (L.) Spreng., and S. reptans (L.) Spreng., and one in the Himalayas, S. elata Royle (alt. 9,000- 15,000 ft.). We have thus before us a genus of quite an extensive geographical distribu- tion, but disconnected and seemingly unexplainable. None of the species are circumpolar; none have been reported from arctic Europe, and nevertheless, two species, oné of which is very characteristic by its stoloniferous habit (S. reptans), inhabit the Alps and Pyrenees. On the Siberian north coast from 173° W. Long. to 68° E. Long. S. glacialis is the only species recorded so far (Kjellman, l.c.); the genus is absent from the Altai and Baikal mountains which otherwise harbour so many arctic species; in the northeastern corner of Asia, however, S. Rossii, S. calthifolia, and S. anemonoides are indigen- ous. But between these regions and the Himalayas no species has been recorded and, as mentioned above, only one occurs in these mountains. On this continent, however, the distribution is fairly well connected. For even if 8. glacialis and SS. calthifolia are very rare on the northwest coast, and none of them crossing the Mackenzie river, we have in S. Rossii a species quite extensively distributed in the arctic region, including the islands of the archi- pelago, and extending south to the alpine summits of the Rockies; furthermore, S. triflora which is not arctic but widely distributed from the Atlantic to the Pacific slope and southwards to Colorado, following the Rocky mountains. Finally, the Alleghenies and the White mountains are the home of S. radiata. But in spite of this fair representation of the genus on this continent, none has been discovered in Greenland. If now the question be asked where these species, or let us say the genus, originated, the answer cannot possibly be in the south, nor in the north alone. The South European elements did certainly not come from the north, since they are endemic to these mountains; they must have originated there, where, turthermore, they are associated with two near allies: Gewm and Dryas. With reference to the arctic species, S. glactalis and S. Rossii, these must have developed in the polar regions, and of these the latter, S. Rossiz, did not altogether leave the south when the arctic flora returned for, as mentioned above, it is still in existence on the alpine summits of the Rockies, in Colorado for instance. S. triflora and S. calthifolia are-evidently of younger origin, both, however, from a northern centre evidently located in Canada. Another centre of development must have been in the Appalachian mountains, so far as con- cerns S. radiata. Finally, with regard to S. elata, so widely secluded from the other species, this must have originated in the Himalayas. Considering these data it seems to be characteristic of the southern element of the genus that so very few species have become developed, and that these are endemic to the southern mountains: S. elata to the Himalayas, S. reptans and S. montana to the Alps and Pyrenees and, finally, S. radiata to the Appal- achian mountains. The reason may be that the centres are of a more recent date, and therefore entirely independent of each other. And it is a point of great importance, I believe, that these independently developed species never- theless show the typical habit of true Sieversia; the only distinction appears in Arctic Plants: Geographical Distribution 107 B S.'reptans where stolons above ground are developed, but otherwise the habit is the same, and the structure of the style, so characteristic of the genus, is identical. And, when a small genus like Sieversia has actually been: produced at stations so remote from each other, but in the alpine regions, there seems to be a corresponding ‘probability to suppose that the same species might also be produced independently from more than but a single centre, as long as the conditions are the same, or at least approximately so, in respect to climate and soil and in regard to association with allied types. _ Ifit were within the scope of this report to include the geographical dis- tribution of alpine species, several instances might be mentioned which would be strongly in favour of Schouw’s hypothesis. However, I cannot abstain from recording one, inasmuch as it is a Canadian plant. I think of the discovery of Papaver pyrenaicum L. in the Rocky mountains: South Kootenay pass, Sheep mountain, Waterton lake, besides that I found some specimens in the U. S. Nat. Herb. from Montana: Near Stanton lake, alt. 7,500 feet; these were labelled P. nudicaule. I might state at the same time that Mr. James M. Macoun sent the specimens to the British Museum in order to have my determin- ation verified, and he was informed by the botanists of the said institution that the specimens proved the first record of the species on this continent. It would certainly be too strange, I think, to explain the presence of this South European species in the Rocky mountains, as being in any way connected with some centre in the Pyrenees. The stations in both countries are alpine, and the association with allied types is about the same: P. alpinum L. and P. nudicaule L. Consequently, it seems probable that P. pyrenaicum originated from two centres: one in the Pyrenees and a second in the Rocky mountains. Returning to the arctic species of the Rosaceae, the genus Potentilla is represented by seven species, two of which, P. palustris and P. nivea, are circum- polar; the others, with the exception of P. fruticosa, are principally arctic types. With respect to P. palustris, this species is nowhere abundant in the arctic region, and it is perhaps one of southern species which accompanied the arctic on their retreat, rather than having actually originated there. P. nivea, on the other hand, is undoubtedly of arctic origin owing to its very wide distribution and frequently abundant occurrence; its occurrence in the Alps, the Pyrenees, Cau- casus, the Altai and Baikal mountains may be explained as it being a remnant of _a post-glacial vegetation left over on these mountains. But concerning P. pulchella, P. rubricaulis, P. Vahliana, and P. emarginata, these appear to be arctic American types, some having extended eastward to Greenland. It is a strange fact that P. pulchella occurs in Spitzbergen although it is entirely absent from arctic Europe; perhaps a former more extensive distribution may have been interrupted. The circumpolar Rubus Chamaemorus is not uncommon in Central Europe but, nevertheless, it is not represented in the Alps of Switzerland nor in the Pyrenees, and it is also absent from the Rocky mountains; in Greenland it is very rare and has only been found between 61° and 64° N. L. With regard to the occurrence of the species in Canada, John Macoun states that it is especially abundant in the north and that, when found at its southern limit, it always occurs in cold peat bogs; as far south as 53° N. L. it has been found in peat bogs at the base of Porcupine mountain in Manitoba. King William island is the only station known from the American archipelago. It is interesting to notice that in Canada it is often accompanied by Rubus arcticus L., as is the case also with the Scandinavian plant. Evidently the habitat “peat bog,’ where the species thrives at its best, explains its absence from the southern mountains, and the centre of its distribu- tion may be sought in the polar regions, even though it seldom becomes sufficiently advanced to produce mature fruit in these regions. 108 B Canadian Arctic Expedition, 1913- 18 Of Rosa acicularis Lindl. a single specimen, a small branch with a well preserved flower, was collected by Rev. I. O. Stringer at the Mackenzie river delta; the species has also been collected on the American coast of Bering strait, and it is recorded by John Macoun from Fort Simpson on the Mackenzie river, Fort Yukon, the Kuskokoin valley, and St. Michael’s island, Alaska, the speci- mens having been collected by Mr. Watson. The species is known, furthermore, from Siberia, viz.: Davuria and Kamtchatka. : Of the Papilionaceae thirteen species were collected by the expedition, exactly one-third of the species known so far to have extended to the arctic region. From Table 1 it will be noticed that three of these are circumpolar, viz.: Astragalus alpinus, Phaca frigida, and Oxytropis campestris. However, their distribution is so extremely scattered within the arctic region that they hardly may be defined as circumpolar; for instance, on the north coast of Siberia, Astragalus has only been reported from Dickson harbour (80° 58’ E. Long.), Phaca only from Preobrascheni island (113° 10’ E. Long.), and Ozytropis only from Dickson harbour. And the other species known from the Siberian coast are just as rare, for instance: Oxytropis Mertensiana only at the mouth of the river Olenek (120° E. Long.), O. nigrescens only at the mouth of the river Kolyma (161° E. Long.), besides Pitlekaj (173° 24’ W. Long.); finally, Pdysaruwm obscurum L. has only been reported from Pitlekaj, and Phaca alpina Wulf. (non Ledebour) was discovered by Sujef on the shore of the Kara sea. To emphasize fully the arctic distribution of these Papilionaceae, Table 5, showing the distribution of all the species of this family within the arctic region, is inserted. It will be seen from this table that, so far, only 39 species have been collected in this region. They represent 12 genera, with 18 species belonging to Oxytropis, 7 to Astragalus, 4 to Vicia, 83 to Hedysarum; the other genera contain only one or two species. 5 a 8 Bg 3 | 8 glee]. 2/5 § part 2 £ 2 |A Tasie 5. = ao| 3 7 = |s o; Fas q 3 8 o | 2 * . * . "1 ont n Geographical distribution of Papilionaceae| a mA] R a a < % 3S A < 3) ii og o Lo] =] g 8 8 in the arctic regions of America, 8 2|Es a a a 3 g ° Europe, and Asia, an ee 31 od 4 ie 5 N a 2-8 28 2 — 3 ‘Bs Bi/ee/fa| 2) E/E) 51 5 | 88 414 < do) am nr =a (ce a 9 | 3 gis i BR bt q q Ee g a = S 8 2 ae < d < < < a oO < iN ia G. Amarella L.......... eae eee! gh dees sa pia eee s . * a) eseareny [Reece caries Lipase eon evasnslereryafe denne Flats eateadiaus = * Vey ses glevadyabnyeuxalegesas G. tenella Ly... 0. eee eee eee ee a ike ‘ * * Me getol| petits a eee * G. propinqua Richards......... EEN ocstee Regsce ell ebdesut tne MW ae otal aga icul] Sah eu beeen haseeianiie * G. arctophila Griseb............ Fe Nha cearetiste | Svante fll issue ip nce | aa acetal ea sawn Pasee acetal wrescenss Sell eopeptorellweied ve _ Se a ee ee ere eee * is FL ostsannyesl aiapeaiei| gaa sien| canes leartartellitstesio a fladvenws G, serrata Gunn................].0.4.. * Oe Maletarei gl alae Acta tiy ate Be Mls debuts sell See caayees OHS Lincerewieveseonaseeaslvanaed * sa Bi gee? dad | Jeet bemeeer arts reer || Paeentieee RPC Ge meena: Licensees aortas eicd gl devine le decors PRO Ries all eeeeye aaa eee ¥ © Wen sins alleamneats G, glauca Pall.............0.0.. hae Sezer | aunndeastelfeeaueenaay * Teccnsyeoaneltauien adele teartene | ste ares * * Furthermore, with respect to the altitudes reached by some of these species in the mountains, the Scandinavian species, except G. serrata, ascend to an altitude of between 3,000 and 4,500 feet according to Blytt (Norges Flora); in the Alps of Switzerland, Heer (l.c.) has recorded G. campestris and G. tenella from 8,500 feet, G. nivalis from 9,000 feet, and G. verna from 10,000 feet; in the Altai mountains G. tenella and G. verna ascend to 6,500 feet, and finally in the Himalayas G. aurea and G. tenella reach an elevation of not less than 10-14,000 feet, according to Hooker (l.c.). _ _ However, several of these are more typically lowland plants occurring for instance on the seashore, notably G. serrata, G. campestris, G. Amarella, and @: aurea. Arctic Plants: Geographical Distribution 121 B In other words, some of these arctic representations of Gentiana occur as annuals or biennials, as alpine or lowland plants, and evidently they all have developed in the temperate regions, G. glauca, G. propinqua, and G. arctophila on this continent, the remaining in Eurasia. However, the occurrence of G. tenella in the Rocky mountains of Colorado is difficult to connect with its wide distribution in Europe unless we presume that the species did originate from two centres, one in the Rocky mountains, and a second in temperate Europe, Scan- dinavia, for instance. And we are confronted with the same difficulty when we attempt to locate the centre for the high-alpine Gentiana frigida Hke. Of this the typical plant is known only from the Carpathian mountains but a variety algida Pall. G. Romanzovit Griseb.) occurs in Siberia: Terra Tschuktchorum, Altai and Baikal, besides on this continent in the Rocky mountains of Colorado and Utah; it is recorded also from St. Paul and Shumagin islands, Alaska. I cannot believe that this species developed from-a single centre, in Eurasia or on this continent. The Carpathian mountains as well as Altai and the Rocky mountains constitute centres of the greatest importance to the distribution and development of Gentiana, and the remoteness of these stations in connection with the association of the species with several allied types may, I believe, indicate its origin from these three mountain ranges. ‘ Pleurogyne carinthiaca Griseb. is a native of the European Alps but it is absent from the Pyrenees; it occurs, however, in Caucasus, Ural, Altai, Baikal, Davuria, Kamtchatka, and the Himalayas (10-13,000 feet); on this continent it has been found in Canada, but only at a very few stations, viz.: Anticosti and Riviére du Loup, Quebec, besides on the arctic coast explored by the expedition, the only arctic stations thus so far recorded. The wide distribution in the south speaks in favour of its actual centre having been located there but it is, of course, impossible to decide whether in Europe or in Asia. The occurrence in North America evidently depends on its distribution from Kamtchatka to the islands of Alaska, rather than suggesting an independent development of the species on this continent where it is, as far as we know, so extremely rare and represented only by specimens of diminutive size and very few-flowered. The other species, P. rotata Griseb., is in Europe only recorded from arctic Russia while in Asia it is known from Altai, Baikal, and Davuria; it is rare on the west coast of Greenland, between Lat. 60° and 69°, while on the American continent it is distributed from Labrador and Hudson bay to the northwest coast of Alaska; it also occurs in the Rocky mountains as far south as Colorado where it is not alpine. It is evidently of southern origin, but of course the actual centre of its distribution cannot be decided. We have thus in the genus Pleuro- gyne two species which are annual and of which the distribution extends across an area of enormous extent throughout the northern hemisphere, but of which P. carinthiaca has reached a wider range in Europe, viz.: to the Alps and Caucasus. It seems strange that the Gentianaceae, though principally high-. alpine, are so scantily represented in the polar regions, and especially by annual or biennial types. The almost exclusively American family Polemoniaceae is in the arctic regions but poorly represented. Polemoniwm boreale is circumpolar, and outside the polar regions it has only been found in the Altai mountains. It is thus absent from the Rocky mountains which otherwise constitute such an important centre of the genus. It would seem very strange if this species were of a more southern origin since, at present, it is almost confined to the arctic region, and especially because, as stated above, it is absent from the Rocky mountains. And with respect to the other species of the genus, the variety villosum is almost circumpolar, while the typical plant is widely distributed in the south, in Europe, Siberia, and on this continent. It actually appears, by considering the geographical distribution of these plants, as if the genus” has (122 8 Canadian Arctic Expedition, 1313-18 originated from more than one centre, yet given rise to species so closely related ° as the two in question: P. caerulewm and P. boreale. The question as to their original centre seems just as complicated as that of Sieversia. . A similar difficulty arises when we consider Phlox Richardsonw, a member of a genus so exceedingly well exemplified in the Rocky mountains, and yet this species is known, so far, only from a very few stations on the arctic American coast. No other explanation seems plausible than the considering of the species as a remnant of the glacial flora, and as the sole remnant of the genus, formerly distributed much farther north but exterminated with but this exception at the high-northern latitudes. : Very few Boraginaceae have been found in the arctic region. Ledebour, for instance, mentions only 1 species of Mertensia, 4 species of Myosotis, 1 species of Eritrichium, and 2 species of Echinospermum as. occurring in arctic Russia, and only 1 species of Myosotis, and 1 of Eritrichium as inhabiting arctic Siberia; no additional species are recorded by Kjellman from the north coast of Siberia. Of these Myosotis silvatica was found on the north coast, but only the typical plant; in the arctic region the variety alpestris Koch is more common and, as described by Ledebour as being “humilior, racemis brevioribus densius incum- benti-pilosis, pedicellis crassioribus, calycibus majoribus,” it is quite distinct from the type. According to Ledebour the typical plant has been found in arctic Russia but nowhere in Siberia, except in the Altai mountains and Davuria. While Mertensia paniculata has been reported from many stations in Canada and the United States, M. Drummondii has never been found outside the north coast where, moreover, it seems to be extremely rare. These two species belong to the section Humertensia which is quite well represented in the Rocky mountains especially, and the occurrence of M. Drummondii solely on the north coast may be explained in the same manner as that of Phlox Rich- ardsonit. But with respect to M. maritima, this is a seashore plant and, although wholly absent from the Siberian coast, it is not infrequent on the arctic coast of this continent, as well as in Greenland and arctic Europe. Being a seashore plant, and at present so widely distributed in Europe and on this continent, the species evidently belongs to the category of plants which originated in the south but gained a more extended distribution toward north when the ice receded, like Halhianthus. ; The large family of Scrophulariaceae is in the arctic region well exemplified by the genus Pedicularis but the other genera are very poorly represented; for instance, according to Ledebour: Limosella, Veronica, Castilleja, Bartsia, Euphrasia, Rhinanthus, and Melampyrum are the only ones, besides Pedicularis, known from arctic Russia; Veronica, Castilleja, and Pedicularis are the only genera known from arctic Siberia. But among these arctic Eurasian genera is one which more properly belongs to the North American flora, viz.: Castilleja. Of this genus C. pallida and the variety septentrionalis Gray, so widely distributed in Canada and in the mountains of the United States, viz.: the alpine summits of the White mountains and Green mountains of New England, and throughout the Rocky mountains, occur, furthermore, in arctic Russia and Siberia, in Ural, Altai.and Baikal, in Davuria and in Kamtchatka, besides in Alaska, and in South Greenland. The geographical centre seems naturally located in the Rocky mountains where the species is associated with several closely related congeners and from where it became distributed to the north, extending east- ward to the American archipelago and Greenland, westward across Bering strait to Siberia and arctic Russia. With respect to Pedicularis, the accompanying table (Table 8) shows the species which have been recorded from the polar regions. This table comprises only 16 species, a small number, indeed, when we bear in mind not only that 120 species are récognized as “‘valid species” by Bentham and Hooker but also, and quite especially, that the arctic element represents sections so exceedingly well Arctic Plants: Geographical Distribution 123 B 8 rr) _ > 2 Tasue 8. 3 3 as 2 < 3 = 3 3 3 8 Geographical distribution of arctic species of} © 4 (Q 6 a S x e Pedicularis. Zz 3 ~ 2 3 g 3 s| co} e/a ]/e] aul gy 2 So) Fe me Be ig | et | a ag < 1) Mm | < < < 4