PROCEEDINGS OF THE Ohio State Academy of Science Volume IV, Part 6 SPECIAL PAPERS, No. 1 1 > THE WILLOWS OF OHIO A MONOGRAPH BY ROBERT F. GRIGGS PUBI .ICATION COMMITTEE : JOHN H. SCHAFFNER JAS. S. HINE E.L.RICE Date of Publication, November 25(b, 1905 PUBLISHED BY THE ACADEMY COLUMBUS, OHK I coli'mbvm: Spam a ■ i .1 p ns. Prin 11 «-,, ^ NOTE. The expense of publication of this monograph is covered by a special grant from the Emerson McMillin Research Fund and a small grant was previously made to assist in meeting the ex- pense incurred during the progress of the investigation on which the paper is based. William R. Lazenby, John H. Schaffner, C. J. Herrick, Trustees^ INTRODUCTION. The purpose of this paper is to make the willows of Ohio knowable to persons of moderate skill in the determination of plants. Perhaps it would be too much even to hope that it will enable beginners in Botany to deal with the willows. But if it helps those who have already some knowledge of the native flora to extend their acquaintance to this very common genius, it will have justified its preparation. For Salix, like Carex and Cra- tegus is considered by many Botanists too difficult for any but the specialist. Many competent workers seem to be unable to cope with Salix and there are few even of the larger herbaria in which the willows are correctly determined. The reason for this the writer believes to be not in any inherent difficulty of the group for it is not so difficult as many better known genera, but in the fact that an adequate description of it has never been presented. It is toward filling this need that the present effort is directed. The species are all rather similar and variable but the dif- ferences between them are not as inconstant as has been supposed. But the space that has been devoted to them in the manuals is entirely too small for their accommodation even when treated by such a master as Bebb, whose account published many years ago in Gray's manual remains the best treatment of the species within our area. Nor can their character be represented by line drawings such as appear in Britton and Brown's Illustrated Flora; even the lithographs of Sargent's Silva are but little better. The charcu ter of a willow leaf is too subtle a thing for the ordinary scientific artist to portray; for that, the camera is necessary. Another fault of most of the descriptions and keys hitherto pub- lished is that they have been written with a complete specimen in mind, as it might be assembled on the herbarium sheet with both kinds of flowers and leaves. But a collector never has a complete specimen and is sure to be balked by lack of the mis- sing parts. It must be admitted, to be sure, that there are some stages of some species which are almost indeterminable. Bui they an- not so numerous as to make n inadvisable to construct keys lor the others. Any successful treatment of the genus musl for the present be local in its scope for Salix is subject to very greal raphical variation and a treatment o\ a given species which would be en tirely accurate in a given area might be entirely inadequate for the satin- species if observed a thousand miles l'min the li- cality. The de cription of Salix n\ ven below for example would not cover at all atisfactorily the southern and western plants winch go by thai name. And yet though the characters 258 OHIO STATE ACADEMY OF SCIENCE. which we use in the descriptions vary enormously, one would still recognize without much difficulty that the southern plant was Salix nigra in a metamorphosed condition. Consequently, as far as the native species are concerned, at least, though it is hoped that this paper will be of service all over the northeastern portion of the United States, one must expect to find it more and more at fault as one recedes from Columbus. Some day when there is a monograph like the present covering every state in the coun- try, some genius will write a general treatment that will effect- ively handle all the species. One reason why it is necessary at present to limit the treat- ment to a small area is that no one has yet succeeded in describ- ing the qualities by which we recognize a willow. We define the species by external, artificial characters such as the shape and hairness of the leaves, the length of the pedicels and so on; but all such characters may vary enormously and still leave the qual- ities by which the expert determines the species, constant. We cannot study willows profitably by tables of dimensions and geometrical descriptions of the leaves any more than we can study the faces of our friends by measuring the length of their noses. As there is a certain almost indefinable individuality in a human face, there is in a species of plants an individuality no less indefinable but no less important. I here desire to extend my thanks to friends who have aided me very greatly in the preparation of this paper: to Dr. W. A. Kellerman of the Ohio State University, who at one time planned to be joint author with me, for innumerable kindnesses done and suggestions given during the whole course of the work; to Prof. John H. Schaffner of the same institution for help especially in verifying and testing the keys and to Mr. Otto E. Jennings of the Carnegie Museum for the loan of the collections of that insti- tution. The present paper is based on studies and collections of wil- lows begun in the spring of 1898 and carried on as opportunity afforded ever since. During that time the author has been enabled to study the plants in the field in various parts of Ohio, in the vicinity of Washington, D. C, and in the Red River Valley of North Dakota and Minnesota. The herbarium on which he has mostly depended is that of the Ohio State University at Colum- bus. The United States National Herbarium and the Herba- rium of the Carnegie Museum at Pittsburg have also contributed considerably to whatever value the work may possess. TERMS. Most of the terms used are those common to Botanical des- criptions, but those referring to the venation may be so unfa- miliar as to require definition. They are based on a most excel- lent paper by Dr. Glatfelter (Rep. Mo. Bot. Gard 5:46-60. 1894.) on the venation of the willows. Primary. A main vein branching directly from the midrib. Costal. A smaller vein from the midrib, of considerable size but not reaching as far as the primaries. ;ondary. A vein of the second order given off from a pri- mary. Tertiary. A vein given off from a secondary. Arch. The joining near the margin of a primary with a fork >cnt off from the one next above, the two branches together forming the arch which rests upon the two primaries. /.(><>/>. Funned by a primary near the margin curving yard and inward and merging into the next above, similar to an arch but lacking the angle at the apex. Regular. Veins parallel. Bract. In his treatment of the genus in Britton's Manual Scale. I Mr. P. A. Rydberg ai >| dies the term bract to the rudimentary perianth of the flower, a small leaf subtending the ■ntial organs, whereas previous writers have termed this a scale ami reserved the former term tor the bracts which support aments. This departure from the accepted usi 'ins to confusing and unjustifiable as the departures from the rules of priority in nomenelature which the New York Botan- condemn so strongly. Further the meaning given the terms in tin- glossary at the end of the book, accords better with the older usage than with Rydberg's application of them. Th definitions arc here- followed. According to Britton, a "Bra is"Aleaf, usually small, subtending a (lower or a flower cluster or a sporange. " That is, in this case one of the leaves which ap- pear at the b >f the aments. And a "Scale" i< "A minute rudimentary or vi leaf." In tin- case one of the small leaves in the axils of which the flowers are b< »rne To make the matter worse, Rydberg uses bract in both ntroducing ambiguity a) The Willows of Ohio. THE GENUS SALIX AND ITS RELATIONSHIPS. The willow family contains but two genera, Salix and Populus. These, in most cases distinct enough in their leaves are separated by the following floral differences. Populus has fimbriate scales, cup shaped discs, elongated stigmas and many stamens, while Salix has entire or only slightly crenate scales, mostly glandular clavate discs, short stigmas and few stamens, two in the majority of the species. In the polyandrous willows which are treated first in the present paper, there is a distinct approach from the diandrous or prevailing willow type toward the poplars, most conspicuously shown in the increased number of stamens but also evidenced in the disc which becomes com- pound with several glands and sometimes almost forms a lop- sided cup. It is, too, in the polyandrous willows and in those diandrous species manifestly most closely related to them that we find most of the arborescent species — a habit prevailing among the poplars — while most of the willows are shrubs. There are recognized in all something like two hundred species of willows. They are mostly natives of the north tem- perate zone but are not entirely absent from the torrid and south temperate zones. The willow may be counted one of the most successful of present day plants. In few other genera, so compact and homo- geneous in respect to their floral characters, is there shown so great a variety of adaptations to varied conditions of life. With- in this one genus may be found plants all the way from large trees to dwarf herbs, affecting habitats from the fertile alluvial plain to the barren mountain peak. They grow almost every- where bill yet there are certain limitations in their habitat. Varying from hydrophytes to xerophytes, they arc uniformly lovers of the sun and never found to any extent in deep tor, • or other shaded situations. In their various habitats their etative structures undergo wide modifications to accommo- date them to their environments. Some species like Salix lucida have broad thick leaves, protected from too severe conditions l>v the hard ;sdossy SUrfai I In others like SdltX adenOphylUx the same protection is gained by ;t thick coating of wool on one or both surfaces. Or a heavy coal of glaucescence may be dev oped probably to ;, degree at least for the same purpose. In other i .1 es tin- leaves are so small as to enable them to endure tin- most severe condition cially when, as is often th< 262 OHIO STATE ACADEMY OF SCIENCE. there is coupled with the small surface exposed a prostrate habit which shelters them from the wind and allows them any ad- vantage to be gained from warmth radiated from the ground. In great contrast to these dwarf creeping almost herbaceous forms are the tree willows so well known to every one. These, by their rapid growth are enabled to succeed well in their favorite habitat along streams and lakes because they can quickly repair the damage done by floods and storms, while more slowly growing trees would be almost hopelessly beaten down and destroyed before they could rear their trunks to a sufficient size to enable them to withstand the elements. Like- wise their well known ability to grow from cuttings and broken branches renders them able to use the very storms which break them in pieces as the means of their further dispersal. Their well known variability under different conditions is an evidence of plasticity of constitution and adaptability to various environments which is as important a factor in the success of a plant in meeting the competition to which it is subjected as is the same quality in the success of a man. From their adapta- bility, variability and the large number of intermediates between very divergent forms we may safely infer that we have here a group of species with a very large amount of "Vital Motion" in rapid course of evolution. ON USING THE KEYS. To construct a key which will hold for all individual willows is well nigh impossible, because in the first place, there is a greater or less number of hybrids for which no pretense at a key is made ; and secondly, there is so much variation in some of the species that even after including them in two or three different places, the writer is not so sanguine as to suppose that he has covered all the variant forms. Bearing this probable defect of the key in mind, it will be understood that the greater the diligence with which each clue given is followed out, the greater will be the probability of a correct determination. If the plant you have seems to fit in both divisions of the key by all means run it through both and after doing so refer to the descriptions and plates as well. The synoptic table is intended primarily to give some idea of the relationships of the different species and consequently the whole anatomy of the plant is used while all the parts are never present in a single specimen. But in the keys proper every- thing is subordinated to the end in view, namely the identifica- tion of the plant ; and only such characters are used as are actually present on a single specimen. In the foot-notes under the keys their especial uses and limitations are discussed. THE WILLOWS OF OHIO. 263 Synoptic Table of Ohio Willows. Stamens 3 or more, filaments pubescent at base, catkins appear- ing on leafv branches. Polyandrae. Trees with rough, flaky, brown bark and brown twigs. Amygdaleneae . Capsules short pedicelled, short globose conic, forming dense thin catkins; leaves green on both sides, glabrous, pri- maries distant, very line net veined, with a marginal running nearly to the base of the leaf, linear-lanceolate when mature, often falcate; of very scragly growth. seldom upright; frequenting streams. 5. nigra. Capsules long pedicelled, short conic, in lax aments; leaves sometimes 15 cm. long, typically oblong with straight sides, glaucous beneath, generally hairy, primaries ascending, but scarcely forming a marginal; sprawling shrub or straggling tree; southern, river banks. S. Longipes. Capsules long pedicelled, long conic, forming large loi catkins; leaves glaucous beneath, glabrous, primaries close, reticulation not so fine, marginal seldom running below the middle of the blade, broadly lanceolate, petiole often red; an elegant shapely tree; northern, ] >rcfcrs swam] is. >. amygdahides. Shrub, twigs shining brown, scales often dentate, capsules large and glossy, aments thick; leaves ovate-lanceolate, often very long attenuate, sub-coriaceous, -lossy above, light green, glands very prominent, especially on stipules; mostly in rocky wet places, northern. S. lucida. Stamens less than .">. Stamens 2. Diandrae. Filaments pubescent, catkins appearing with the leaves, on leafy branches, except early flowers of >'. interior, ales nearly as long as the ovularies at anthesis. Large trees with barkroughbuf not flaky, leaves glaucous below, capsules glabrous, short pedicelled. Fra Not weeping, primaries and secondaries close and regular. Bark greenish, leaves glabrous, rather coar er rate, and trongly glandular; capsules long i nt -, remaining green. S. i igilis. Bark yellowish, leaves pubescent at least when young, fine serrate; cap ules short conic, turning yellow m frail S. al hark golden yellow, leaves glabrous in a. S alba viU .'. 264 OHIO STATE ACADEMY OF SCIENCE. Branches pendent, leaves small, very fine net veined, primaries and secondaries irregular; capsules short globose conic, remaining green in fruit, staminate plants practically unknown. S. babylonica. Shrub, often growing in dense clumps with many slender stems; leaves often very long, linear-oblong, with a strong marginal vein and distant primaries, smaller veins except a few costals vanishing; catkins often in cymose clusters which continue flowering all sum- mer, or the earliest only leafy bracted. 5. interior. Leaves short and broad (more than 1 cm.) inclined to be very wooly ; cymose clusters of catkins very pro- nounced, carpellate plant rare. 5. interior wheeleri. Filaments glabrous, catkins coming before or with the leaves but not on leafy branches unless in fruit. Capsules pubescent at least in flower. Styles short and inconspicuous, less than half as long as ovulary. Leaves mostly broad, coarsely serrate or entire; cap- sules long conic, very hairy ; upright, many stemmed shrubs preferring lowland swamps. Capreae. Catkins at anthesis seldom 2 cm. long, not very wooly, bracts conspicuous, scales light brown, persistent, pedicel very long, filaments slender; venation prominent below, leaves dirty white tomentose, especially on the veins. 5. bebbiana. Catkins seldom less than 2 cm. long, wooly pussies, bracts small, scales almost black, filaments thick, pedicel medium; leaves mostly glab res- cent, veins not very prominent below. 5. discolor. Leaves narrow. Leaves undulate-revolute to entire, primaries prom- inent on the under surface, distant, coming out at right angles and arching or looping regularly to the base of the leaf ; aments small short pussies without leafy bracts, capsules long conic, very hairy; upland swamps and hillsides. 5. humilis. As above but smaller throughout; leaves up to 5 cm. long, aments less than 1 cm. long, shrub less than 1 m. tall; prairies especially. 5. humilis tristis. THE WILLOWS OF OHIO. 265 Leaves sharply serrate, showing a decided tendency to blacken in drying, aments from pussies, bracts small or none, capsules when long conic, thinly pubescent; lowland shrubs. Sericeae. Capsules blunt, short pedicelled, short conic, densely silvery silky ; leaves dull above, lustrous sericeate below, at least till old. 5. sericea. Capsule long conic, long pedicelled, sparsely pu- bescent; leaves shining above, glaucous be- low, glabrous (or rusty sericeate when young); northern. S. petiolaris. Style very long, capsule short conic, silky, white; leaves long and narrow, revolute with veins prominent below and depressed above, not distinctly arching, snowy tomentose below; wet prairie shrub not more than a meter tall; northern. 5. Candida. Capsules glabrous, styles short, filaments glabrous. Filaments distinct to the base; leaves sharply serrate, good sized shrubs. Cordatae. Leaves dull on both sides, only a little paler beneath Leaves thickly pubescent on both sides alike, short and broad, bracts broad, remaining green; north- ern. 5. adenophylla. Leaves thinly pubescent with most of the hair on the veins beneath, or glabrous, green on both sides or slightly glaucous beneath, lanceolate, bracts narrow, remaining green. S. cor data. Leaves -lossy above, very glaucous below, glabrous, generally broad, bracts narrow; turning black; northern. >'. glaucophylla Filaments frequently united a1 the base, capsule long pedicelled in fruit, leaves small, entire, purplish een, conspicuously reticulated; low shrub in i.ignuni bogs, S. pedicellaris. Stamen I, formed by the coalescence of 2, anthers i. scales black, reflexed in staminate flower, capsules rarely seen, hort , thick, hairy ; leaves oblanceolate, opposite or scattered, purplish; shrub with long slender branche Synandrae. S. purpurea. 266 OHIO STATE ACADEMY OF SCIENCE. Key Based on Carpellate Aments.* 1. Flowers appearing in pussies before the leaves, bracts small or absent. 2. 1. Aments and leaves unfolding together, bracts mostly large. 11. 2. Capsules glabrous from a wooly rachis, inner membrane of bud scale growing out beyond the outer. 3. 2. Capsules silky till after breaking open, bud scale not as above. 5. 3. Capsules turning brown when ripe, 10 mm. long, aments lax, bracts glabrous and glaucous in fruit. 5. glaucophylla. 3. Capsules remaining green, less than 10 mm. long, bracts green on both sides. 4. 4. Bracts narrow, obscurely serrate or entire, aments be- coming rather lax, leaves becoming glabrous above S. cordata. 4. Bracts broad, prominently glandular, aments dense, leaves becoming densely tomentose above; north- ern. 5. adenophylla. 5. Style very long and slender, rose red at anthesis; conspicu- ous, capsule snowy white; leaves tomentose; dwarf shrub; northern. S. Candida. 5. Style less than half as long as ovulary, leaves glabrous above. 6. 6. Capsules short conic. 7. 6. Capsules long conic. 8. 7. Aments long, 35 mm. or more, very dense, often opposite, capsules large, strictly sessile, carpellate plant rare. 5. purpurea. 7. Aments short, 30 mm. or less, moderately dense, not oppo- site, capsules small, short pedicelled. 5. sericea. 8. Capsules short pedicelled. 9. 8. Capsules long pedicelled; northern. 10. *The fruiting aments of our willows are quite distinctive and though the differences between the species are not always easy to describe, they are constant and easily learned, so that with fruiting specimens we should have a minimum difficulty in determination. But in flower they are by tio means so easy to determine. The ovularies are all very much alike and seem not to develop character until filled out. In fruit too, the leaves of most specimens are present and help greatly. The greatest need of caution in using the key is to be sure that the descriptions of capsules are not applied to younger stages which are more slender with shorter pedicels. It will also be found difficult to use the key after the capsules have burst. The leaves referred to are those pres- ent at flowering and fruiting time. They mav or mav not be similar to the mature leaves of the species. THE WILLOWS OF OHIO. 261 *). Aments long in fruit with many capsules, stigmas greenish white when fresh. S. discolor. 9. Aments short and thick, few flowered, stigmas red when fresh. S. humilis. 10. Aments short, scales frequently darkened at the tip, leaves narrow, turning black. S. petiolaris. 10. Aments long, scales yellow, leaves broader, strongly veined, remaining green. S. bebbiana. 11. Capsules hairy. 12. 11. Capsules glabrous. 16. 12. Style more than 1 mm, long. 5. Candida. 12. Style less than 1 mm. long. 13. 13. Capsule subsessile, blunt pointed, stigmas sessile. S. interior. 13. Capsule with a distinct pedicel, acute, narrow conic, stigmas rarely sessile. 14. 14. Capsule thinly hairy, almost glabrescent, aments short, leaves narrow, silvery, blackening; northern. S. petiolaris. 14. Capsule persistently hairy, aments long. 15. 15. Scales persistent, light colored, leaves broad, wooly below with prominent veins, pedicels very long; northern. 5. bebbiana. 15. Scales usually deciduous at length, almost black, leaves without prominent veins, pedicel short to medium. 5. discolor. 16. Capsules long pedicelled. 17. 16. Capsules short pedicelled or sessile. 1!». 17. Aments short, bracts obtusish, small, capsul nerally redening; small shrub in sphagnum bogs only. >'. pedi cllaris. 17. Aments long, bracts large; good-sized shrubs or tires. Is. 18. Capsules narrow conic; northern. 5. amygdaloides. 18. Capsules mostly globose conic; extreme south. S. longipes. 19. Capsules short conic, aments nearly glabrous. 20. 19. Capsules rather long conic 22. 20. Aments d< i hort (25 mm. or less) of small calibre, with a few small bracts. S. babyloti l'O. Anient er, bracts larger. 21. 21. Scales as long as the ovulary, capsules yellowing in fruit, aments rather lax ; leaves pubescent . glaucescent S .///»'. discolor. 12. Pubescent on both surfaces. 13. 12. Pubescent only below. 1 1. 13. Leaves ovate, subcoriaceous, very glandular, rarely entire. tomento e with matted wool, green on both side> ; shrub, northern. S. adenophylla 13. Leaves lanceolate, thin, nol tally glandular, pubescent with parallel hairs, glauo >us beneath ; t <• .■//>. hi 39. 38. I. dull above, no1 t alwa with a trace ol hair on the veins above or bel< M" ' petiole touti h. xta. 272 OHIO STATE ACADEMY OF SCIENCE. 38. Leaves subglossy, frequently very glaucous, when young sometimes pubescent but with evenly dis- tributed sericeate hairs, inclined to be distantly ser- rate with prominent glands, petiole slender, leaves hanging down gracefully from the twigs. 5. petiolaris. 39. Venation mostly regular, reticulations not very fine, no marginal; trees with smooth bark on the branches. 40. 39, Venation mostly irregular, reticulations very fine, with a strong tendency toward a marginal ; bark of branches ridgey, brown. 41. 40. Bark of branches yellow to olive; serrations very fine and sharp, not especially glandular. 5. alba. 40. Bark of branches green; serrations rather distant, glands prominent, glauscescence absent from the veins beneath leaving them contrasted with the rest of the leaf. S. jragilis. 41. Northern and western parts of state only, shapely tree, especially in swamps; leaves ovate-lanceolate, twigs smooth. S. amy gdalo ides. 41, Extreme south only, straggling tree or sprawling shrub in river beds; leaves oblong-lanceolate, twigs usually pubescent. 5. longipes. AMYGDALENAE, BLACK WILLOWS. Mostly trees with rough brown bark, leaves long attenuate, often falcate, with a marginal vein at the tips at least, reticula- tions of the secondaries and tertiaries very fine. Catkins un- folding with the leaves, stamens 5 or more, filaments pubescent, scales one colored, deciduous, capsules glabrous, green, style short or none. Conspicuous inhabitants of the temperate re- gions of America and extending clear through the tropics to Chili and Peru. For a key to this section of the genus I feel that I cannot do better than copy the following table in which the characters of the three species are contrasted, from an article by Dr. Glatfelter in Science Nov. 1, 1895, which I have found quite useful. It is constructed for the forms about St. Louis and is particularly servicable on that account ; for Missouri is the only region where the ranges of all three species overlap and all can be studied together. THE WILLOWS OF OHIO. 273 5. nigra Range extended north and south large tree, branches crooked, ascend- ing stems in clumps old bark flaky twigs very brittle at base hardy shoots pubescent leaves oblong or lin- ear lanceolate bases acute to trun- cate under surface green venation very min- ute, strong mar- ginal petiole short stipules pointed, persistent, non- glandular blossoms about Ap- ril 25 stamens mostly less than 6 scale- short , obtuse ca] >sules ov< rid coni- cal pedicel short . slender notched stigmas and style S. longipes south of 39° lat. 5. amygdaloides north and west small spreading top large, branches single deeply laticed ridg'y tenacious buds winter-killed hoary pubescent same or broader straight single smooth or roughish somewhat brittle hardy glabrous ovate-lanceolate acute to auriculate acute to cordate whitish glaucous bluish glaucous without marginal short obtuse, persistent, non glandular May 5 1 7 mostly 5 6 ovate globose conical long, stout both poorly devel- oped. coarser more regular very long obtuse, caducous, al- ways glandular April 1") 6 '.i ovate-oblong, acute ovoid conical long, stoutish as m nigra. SALIX NIGRA Marsh. BLACK WlLLOW. A tree occasionally reaching a height of lit meters and a trunk diameter of 1 meter but generally dying when about 15 meter tall, h most often grows in characteristic straggling clumps with lour or five i I, leai ems. In the tall it drops off most of th< i ; the old branches bare, a habil which hinders their elongation and in a tew makes them thick and stubby. Tl oi elf-pruning together with the i raggling clumps in \\ hich it gri iws give it a habit which n typical specimens clearly distingui it from any other Plate I. Salix nigra. THE WILLOWS OF OHIO. 275 species, especially in winter when it is leafless. In summer the long slender twigs sometimes give it almost the appearance of the Weeping Willow from which, however, it can be easily distin- guished by its leaves green, not glaucous The winter buds are very small, le - than 3 mm. long, broadly ovate, acute, commonly but not always without the mark of the leaf base across the back. The leaves commonly are about 10 em. long by 1 wide, narrowly lanceolate with a very long attenuate-falcate tip, mostlv very finely serrate, green and glabrous on both sides. The character- istic venation of the Amygdalenae is at its fullest development in Salix nigra. The marginal vein often runs almost to the very base and the secondaries and tertiaries blend into a system of meshes tiner than in any other of our species. Sometimes broad, blunt leaves are found at the bases of lateral twigs but they are still easily recognized by their tine re- ticulation. The aments appear with the leaves, capsules glab- rous, short conic, short pedicelled, forming thin close cylindric catkins which, supported as they arc by the characteristic leaves, mble those of no other species. The stamim emble very closely those of >'. amygdaloides (which - - stamens 5 or more, filaments pubescent. Everywhere throughout the state, Salix nigra is our com- le t willow. Hut it assumes a much more important role in plant society along the southern border than further north. There it attains its greatesl size and at the same time becomes nnnh more abundant than elsewhere. For long distances along the Ohio River it is almosl the only native willow met with and occupies all the territory wlueh in the north is divided up between era! species. So far as Ohio i- concerned it is perhaps the most constant and easih nized of our willow, hut in the south it is aim ntical with S. humboldtiana and is rather hard to separate from S. longipes. In the west it i represented by several variable varieties which seem to conned it with related forms and render it ;i very difficult subject indeed. I' typical habitat is alone i reams hut it may he found in wet pi;. nerally though it seems to prefer moving I ignant water and is much more infrequent in swamps. The variet; "falcata" is a form with narrower more falcate leaves. In my opinion it i cely worth ideration since ,t is no1 etically different but is merelj an accidental leal variation without correlated variation in other char el i • • ■ L( ■ inary growth and of rai iwth with stipules; flov and fruit typical; natural ize; di of th< flowers and ca| i i .-mi. ra lui ida md | 276 OHIO STATE ACADEMY OF SCIENCE. of the plant. All the leaves are more or less falcate ; their falca- tion varies greatly on the same plant. A hybrid between 5. nigra and 5. alba has been reported from New York.* It is said to have the catkins of 5. alba and the leaves of 5. nigra x amygdaloides . Such a cross is most surprising in view of the distant relationship of the parents. It has not been reported from Ohio. Salix loxgipes Shuttlew. Ward's Willow. This species has not as yet been found in the state ; its near- est reported station is at the falls of the Ohio at Louisville. Though this is some distance south of our territory it is possible that it may be found along the Ohio River. I include it here be- cause of that possibility and in order to make the paper more useful outside the state. Its range extends from Missouri to Washington, D. C, and southward to the Gulf. Dr. Olatfelter says that it is not found like 5. nigra sometimes away frcm the banks of the streams but is strictly confined to them. Sometimes it grows into a tree like 5. nigra but much more bushy. Around Washington it is a low shrub reembling 5. cordata surprisingly, considering the remotene r; < f their relation- ships. The leaves have short stout petioles which with the mid- ribs and larger veins are usually hairy. The blades are extreme- ly variable. Frequently in rank growth, they are auriculate at the base; this character when present segregates them at once from any other of our species. Sometimes they are very long, oblong-lanceolate with straight edges narrowing gradually to the tip; and this again is like no other of our species. More often they are lanceolate with upper surfaces varying from shiny and glabrous to dull and hairy. Sometimes they resemble those of vS. nigra closely except for the glaucous under surface. Or they may be so similar to those of 5. cordata as to deceive even the expert ; often this resemblance is especially well borne out by the under surface which is at times gray glaucous and hairy exactly like that species when grown in a dry place. The two can best be distinguished by the venation which is similar to that cf the rest of the Amygdaleneae except that the marginal is hardly ] er- ceptible and in its place the primaries arcend a long way near the margin. From printed descriptions Salix amygdaloides might be con- fused with the present species but they are net very similar. Salix amygdaloides is a much cleaner more shapely tree, never shrubby; its leaves are long petioled, decidedly broader, much * Bebb, Abrorcscent Willows of North America 3. Gard. & For. 8:42.> 1895. Fig. 58. THE WILLOWS OF OHIO. 277 more sharply pointed and never pubescent. While both have a bloom on the under surfaces, S. longipes is gray glaucous and S. amygdaloides bluish glaucous. Besides all this their ranges ill i tv »t i overlap in Ohio. In fruit it is easily distinguishable from either of the other Amygdaleneae. The capsules are similar to those of S. nigra but larger and long pedieelled like those of 5. amygdaloides. !.IX AMYGDALOIDES Anders. PEACH-LEAVED WlLLOW. ilix amygdaloides grows to medium sized or rarely to a large tree. Its bark and general appearance suggest at once its affinity for the black willow of which it was once considered a variety. Its habit, however, differs very considerably from •that of Salix nigra. It is generally sinj timed and very shapely, with clean branches and darker brown, smoother bark. The winter buds are nearly twice as lar iliose of Salix nigra, dark brown above with a much lighter base where they were pro- by the the old leaf in the fall. Salix amygda- loides is well-named for when in leaf the tree, at a little distance, often bears a stnki mblance to a peach tree; the twigs and petioles are often reddened and the leaf arrangement is similar to that of a peach tree. Both twigs and leaves are entirely devoid of hairs while young sh t, of the other black willows are pu -it. The leaves are much broader than in the other members the amygdalenae, ig almost ovate, distinctly broad >W the middle with a rounded base and an attenuate fall tip, bright green ab( neath. The venation while of the black willow ty; is to approach the regular type, ex- emplified by S. (///>; twigs shining orange brown, glabrous; buds rather narrow- ly ovate, large, (5-10 '.11111 bright reddi h brown in spri duller earlier in the Leave reaching ai me len oi 18 < in. and a breadth of 8 < m., varying from ovate tolance< late, rounded 1 1 1 at t! 'aildul. te. cially on the di 1 tipule , often cov ed when young with Ion'.' tawnj seal tered h; irs, bi labrou • the plant a very beautiful appi ance differenl fn m an) '.the- willow. The thickness oi the leaf make, the rath. 'ill to make -nit. Stann Plate 1 1 Salt \ amygdaloid All part . typical, natural ale which is enlai tin,-, til Plate III. $alix lucida THE WILLOWS OF OHIO. 281 nate catkins borne with the leaves; easily recognizable by their large diameter and fine appearance; scales large, conspicuous, cre- nate or dentate, stamens mostly 3-5, filaments pubescent at the base. Carpellate catkins also large, dense, 5-8 cm. long in fruit, long persistent, scales obovate, slightly pubescent, pedicel near- ly half as long as the capsule, stigmas large thick, deeply notched, style short, capsule narrowly cylindric, about 6 mm. long. Salix lucida is a northern plant oceuring in the northern third of the state but not extending to Columbus. Salix lucida hybridises with Salix alba and S. fragilis and when all three come together in one plant it creates very gnat confusion in a group already very difficult to handle. Though hybrids are mostly individual ca es and'irreducible to any general type, it may be said that h S. lucida and one of the fragiles are likely to have e of the ing characteristics. Nearly always the lea ■ dull ii y, sometin they are pubescent like 5. alba. They are likely to retain some- pr - ida. Though they may be her hey ai have the large reddish brown winter buds of 5. lut ida. ■ likely t > be . hi thick ; ; tidra L. Burop Shining Willow. ilix pentandra, the I da, ha in the ; ntly cultr i ba k 1 with any S. lucida the na1 told ap Salix pentandra, ' uate lea • t our American form; tin- single bn >ader sembles in' ire i .hi i' irm . li. ii ural 284 OHIO STATE ACADEMY OF SCIENCE. Leaves sharply serrate, pubescent at least when young, glaucescence bluish. S. alba. Leaves distantly blunt-serrate, glabrous, glaucescence greenish. 5. fragilis. From capsules. Capsules long conic, short pedicelled, green or black in dry- ing. 5. fragilis. Capsules short, ovoid conic, pedicel very short or none. Capsules yellowing, catkins often long. 5. alba. Capsules green, catkins short. 5. babylonica. Salix fragilis. L. Crack Willow. . A tree reaching a height of 25 m. and a trunk diameter of 2.1 m. When in clumps it forms a tall slender tree, but a single individual growing alone branches out so that the head is as thick as high. The bark of the trunk is roughish, gray, that of the smaller branches green and of the twigs sometimes red, winter buds large (7 mm.) seldom well filled out. Leaves reaching a length of 17.5 cm. and a breadth of 4 cm.; acute, narrowed to the base, coarsely serrate with prominent glands, glabrous, greenish glaucous beneath, petiole short, stipules early fugacious, venation regular, showing through the glaucescence as a dark net work, primaries ascending, straight, close together, not arch- ing. The aments being accompanied by leaves are easily identi- fied by the leaf characters; stamens 2, with pubescent filaments, ripe capsules 4ong conic, green, with a short but distinct pedicel. Salix fragilis is a European species planted extensively along streams to hold the earth or to act as a wind-break. It is also pollarded for its twigs which are valuable in basket making. It is one of the most abundant species in our area. It is found both planted and escaped everywhere. Our American plant is differ- ent from the typical European form in having narrower leaves but it is in most cases sufficiently distinct from 5. alba. Because of its quick rank growth, shapely habit, and beau- tiful gray green foliage which turns over very prettily with every breeze Salix fragilis is the best of the willows to plant for a shade tree. It is difficult to understand why it has not supplanted Salix alba and its varieties long ago. But the nurseries nearly always carry a larger stock of the latter sorts which are much in- ferior because of their habit of sending out suckers all over their trunks. THE WILLOWS OF OHIO. 285 Salix alba L. White Willow. A tree attaining a maximum height of 30 m. with a trunk diameter of 2."> m. Like Salix fragilis often growing in clumps but when single-stemmed it is taller and more slender and has the trunk continued as a central shaft to near the top. It is not clean like that species hut is covered with a brush of suckers. Bark of the twigs and branches yellowish green varying to yellow ; winter buds smaller (4 mm. long) than those of the crack willow, oblong and well filled out. The leaves reach a length of 13 cm. and a breadth of ;5 em., lanceolate, acute, narrowed to the base, closely and finely serrate, sometimes almost entire, grayish or bluish glaucous, pubescent on both surfaces at least till mature, (hair mostly persistent below) with close, line, appressed, parallel, gray hairs, stipules deciduous; primary veins close (closer than in ragilis), straight, ascending, regular, extending to the margin without branching, secondaries conspicuously regular but often forking like the letter Y. Catkins on lateral branches or some- times supported only by bracts, scales hirsute, deciduous, cap- sules ovate-conic, not more than 4 mm. long, greenish yellow in fruit, obtuse, glabrous, pedicel very short, style short, stigmas thick. Salix alba is a European species planted in this country for the same purposes as S. fragilis. In most parts of the state it does not seem to escape so readily as that species and hence is not quite as common but may be found planted almost anywhere. A- stated above, most observers have considerable difficulty in separating this s\ from the preceding. The difficulty is often assigned to their hybridising propensities. Hut in Ohio at least hybrids are rather rare. I have found that the two distinct and separable m nearly all cases though it was only after long study that the ability to distinguish them was acquired. The manuals state that the species in the typical form is rare in this country, the majority of the American tonus being the golden er (var. vitellina). It i- certainl; true that few of our plants are the typical hairy planl of Linneus but study of the European mat en a I at Washington leads me to the conclusion that in Europe t he- typical form is aboul a- scarce as here. Further the extremi ly brighl yellow twigs and glabrous, half shiny leaves of the typi- cal varietal form are scarcer in this country than the pure alba form-. [1 ood practice in a oi intergrading forms to draw the hue between specie-, and variety dose to the variety and to call all hut nearly typical varietal forms the spei , it lor no other reason to avoid the use of a trinomial. The variety vitellina then, of Salix <»//>./ includes those plai I with brighl golden yellow twigs and branches, and leaves soon glabrous and brighl green. Plate V. Salix alba. THE WILLOWS OF OHIO. 281 The blue willow, Salix alba coerulea, has not yet been recog- nized in Ohio. Indeed there are some willow students of high authority who do not distinguish it at all in this country whatever may be its status in the old world. As hinted above Salix alba hybridises with Salix fragilis though not so frequently as might be supposed. It al with S. lucida as described under that species. Salix babylonica L. Weeping Willow. The Weeping Willow -rows into a large graceful tree 20 25 m. tall, easily recognized by its very long drooping twigs, which have a habit peculiar to themselves of -ending out numer short branches at a very acute angle with the mam stem. Leaves commonly rather small, 7-10 cm. long, about 1 cm. broad, nar- rowly lanceolate, long acuminate and often fa! 1 the tip, narrowed to the base, sharply serrate, glabrous unless very young, greenish glaucous or at least paler beneath, petiole sh stipules apparently absent; primary veins forming regular acute loops which run together into a more or less straight marginal" this together with the very fine reticulations caused by the rela- tive prominence of the tertiaries often gives the leaf a more or less cli semblance to that of S. nigra, which, however, is newer glaucous as in the present species. Aments on rather sh< >rt leaved peduncles, rather dense, not more than '.', cm, long. capsule short conic, glabrous, green, all but sessile, style short. h is a remarkable fact that the staminate plant of this species is unknown in America. It i etimes that it does not ur at all and it i^ sufficiently rare to warrant such an ion but yet in the national herbarium is what 1 believe to be a gen- uim en of tin- staminate flowers. It was collected by ille at [thaca, X. Y., in 1885 (?)• The leaves an' similar to tho rimonly appearing with the carpellate catkin - aments an- shorl than -~> mm. lonj ely flowered with a rhachi ered \\ i1 li hairs. Tl ub- tended by a very short ovate scale which is much shorter than in any other of the Fl Doubl with the Weeping Will with the l m Willow thai the ab I of fli > ents the i ural spread ol tl.< Me for th< few plant 5 ha aped w Inn cond favorable. I'm whateverthi tain that the i hough commonly planted, i 1 during I he Plate V. Salt I. i.e. es and if photi »graphed and bri il Plate VI. Salix longipes and Salix babylonica. THE WILLOWS OF OHIO. 289 seven years I have been continually on the lookout for it but have seen less than half a dozen individuals which were not clearly planted. One of these was on the Hocking River n< Sugar Grove; the others were along the lake shore in Ashtabula count}'. In Europe Salix babylonica hybridises freely with 5. fragilis. But in this country the manuals have not included such a cross. A single plant was discovered at Sandusky during the season of 1903. * The leaves and habit were so exactly intermediate be- ween the two that there could be no doubt of its identity. At Ashtabula was found a plant which from leaf and habit I take to be a hybrid between the present species and Salix alba, which is not reported in the manuals. LONGIFOLIAE, THE LONG-LEAVED WILLOWS. The longifoliae comprise a very distinct and compact group of American willows. They have no clo nities with any other group and do not intermingle with any. They have two stamens (a specimen in the Ohio herbarium has three) and light one-colored deciduous scales which show their relationship with Loth the polyandrous and diandrous willows. Within the group many described species are difncuH nize; Bebb, him- self, said after he had describ o or three of them that he did not know but what they were all one sp< fter all. The group i-, very easily recognized by the venation of the leaves which is different from any other willow and much resembles that of many ■cample the tire weed. There i^ typically a prominent marginal vein running clear round the leaf, connected with the midrib by . of distant nearly straight primaries be- en which there are practically n ind no mesh- work, onl; -tab running parallel to the primaries. B in young leaves the vi i rid at a much sharp< le and the marginal vein i prominent while the secondaries and tertiarii faded from vi< that the above descrip- tion will not hold. The lea long and som< narrow that it i^ difficult to find any veins 11. * Ohio Nal 1:13 1 1903 VI. Salix longipt ■ (left) and s.;/.'> babylot l • tens lKiim • hree tii Plate VII. Salix interior and Salix interior wheeleri. THE WILLOWS OF OHIO. 291 Key. Leaves narrow, not over 1 cm., sometimes very long, second- ary, (auxilliary) aments very much younger than the primary; forming dense clumps of slender stemmed shrubs. S. interior. Leaves often 1.5 cm. broad or more, not long in proportion, secondary aments of about the same aye as the primary, forming a cymose cluster; low, bushy, not in close clumps; northern. S. interior wheeleri. Salix interior Rowlee. Long-leaved Willow, Sandbar Willow. (S. longifolia and S. iiuviatilis of the manuals in part.) The characteristic habit of this plant is to grow in clumps about 4 meter- high with a great many slender stems coming up close together from a common root system. Th< ms more slender than those of any other of our willows and when the species assumes this habit it may lie recognized msiderable distance. Unfortunately, however, it di not always do so, but sometimes grows by itself as a bush or small tree and then it can besl be distinguished by its leaves. 'I twigs : ",- slender, thickly branched and straight ascending so that the stems hav< eculiar brush-like appearance which gives ;i pleasing softness to the landscape. The winter buds are small and thi I ible th the Mack willow more or • a- the bark is of about the same color. Main- of the bud iff early in the autumn. The pla< uch are taken by small lateral buds which develop one on each Id scar. This habit when present is characteristic. It i ponsible for the number of small hes which out for there the < <. twigs instead of one at Th imetim i l 5 cm. ) and no1 thanlcm.wid dth throughout, Lr-obloi but with shallow distant spinu- th. When they s The venation is ; m young leav< ribed ab with matted until very old, bu1 glabrous and i hey unfi »ld. and Sat Leaf) branch of 1 1 1 « - i >rm of S fruiting amenl shown, i he other with i three ti mi \ the dei « hich com i he the 292 OHIO STATE ACADEMY OF SCIENCE. The flowering season of Salix interior is much longer than that of any other of our willows; it begins just after the pussy willows have gone by and continues late into the summer. I have even seen blossoms in October and on Cedar Point they are not uncommon in July and August. The first catkins come out on short peduncles with a few small bracts. Later when the season's twigs have developed, they also bear aments at their tips. Just below these terminal catkins develop other lateral a- ments which blossom later and so prolong the season. We have no other willow which does this and the presence of these small undeveloped aments is very characteristic. The carpellate aments are generally but not always quite lax as they grow older. The flowers have a tendency, sometimes very marked, to appear in fascicles of from three to five on the rachis with a distinct in- terval between them. This is another characteristic feature present in no other species. The scales are yellow, deciduous, the filaments frequently pubescent. The ovularies at anthesis are scarcely longer than the scales with sessile stigmas on their summits. They vary much in shape being sometimes, especially when very hairy, thick and short with a squarely cut off tip, sometimes nearly rostrate especially when glabrous. The ma- ture capsules are narrowly conic, blunt pointed so as to be almost cylindric if not well fertilised. When well developed they are quite large (1 cm.) sometimes glabrous sometimes tomentose. This variation makes them a puzzling problem and one would sup- pose there were several species instead of one but there seem to be no lines of cleavage between the different forms. Salix interior is with the exception of its own variety wheeler i and the Texan 5. thurberi the only representative of the longifo- liae east of the Rockies. It extends all oves the Mississippi valley and is occasionally met with east of the Alleghanies. In Ohio it is common everv where. Salix interior var. wheeleri Rowlee. Wheeler's Long-leaved Willow. This variety as I have seen it in Ohio sometimes acquires a slender tree form, but more generally is a low much branched dwarf bush, spreading in the sand by the sprouting of buried stems. These do not as in the species produce a dense clump of stems close together but come up only at distances of a meter or so and the result is a loose clump the members of which appear like independent plants. In extreme forms, the leaves especially the older ones from the axils of which branches come out, are very much broader than in the species (7-10 cm. x 2 cm.), dark green and glabrous with the typical venation of the longifoliae, except that the primaries are rather closer and more ascending. These extreme forms as they intergrade into the narrow glabres- cent leaves of the species pass through a series of forms which THE WILLOWS OF OHIO. 1".':'. are somewhat narrower but so extremely wooly that neither sur- face of the leaf tan be seen. It was from these that Prof. Rowlee took the type of the variety. These intermediate leaves shade into narr< nentose forms which < t with the narrow glabrous 1 of the typical species. One might doubt, if he tudied the plant in the field, that the broad glabn leaves were the extreme form, were they not accompanied by an extreme in floral development which is more significant than the leaf characters on which Prof. Rowl< the variety. This is in the development of the secondary aments at th of the terminal. In the variety the aments often form clusl of half a . all opening at nearly the same time — a thing which is rarely seen in S. interior itself. The flowering period is also distinctly later in the variety than in the species. ( )n Point it seems to be at its height the first of July and continues into August. Fully 99 ' , of the plants on Cedar Point are stam- To what the dearth of carpellate plants is i\u< sible to say at present. As yet the vari 'heeleri has been found in Ohio only at Painesville (II. C. Beardslee no. 117, fide Rowlei lar Point where it is abundant. It possibly occurs all along the lake ire and possibly for some distance back into tin- country. B I was unable to find it in Ashtabula county though the conditi' :n favorable. Its author limits its ra of the . CAPREAE. Low I ir shrubs with leaves ordinarily broad in propor- tion to their length, rally glabrescent above, mostly tomen- ath, catkins appearing very early, oftenest m ; in our illi ius. Key. From lea\ Leave ordinarily very tomentose beloM on the rugose veins, venation strongly sunken above, northern. S ina. Li ten glal lised < >n the under nor depressed above, when hair with red-brown haii From fl Catkii with the i »r i ml; a lit tie while fore them, small at anthe only al the tip, narrov lindric, filaments not coarse n< Salix bebbiaxa Sarg. Bebb's Willow. A shrub or small tree occasionally reaching a height of 8 meters, with a habit almost exactly like that of S. discolor and conspicuously different from that of all of the other shrub wil- lows in that there is scarcely any tendency to sprawl, but the stems all ascend from the root. Leaves generally elliptical, varying from sharply serrate through undulate-serrate to entire or often slightly revolute, generally glabrescent above, wooly be- low at least on the veins, primaries and secondaries prominently raised on the under surface making them very conspicuous, pri- maries rather distant, inclined to be crooked and often forking. The whole system of veins strongly sunken from above. Catkins appearing just before or with the leaves, with leafy bracts or, in fruit, on leafy branches; staminate 3.5 cm. long or less; carpel- late sometimes b' cm. in fruit; scales yellow or slightly darkened at the tip, pubescent, persistent in fruit; capsule long pedicelled, villous with white silky hair, cylindric, obtuse, sometimes 1 1 mm. 1( mg in fruit. Salix bebbiana is found across the northern third of the state. This species in its normal forms is very distinct from Salix discolor and ran be separated from it without the least difficulty but the western forms though most keep their flowers like the type, have lea resembling tho e of S. discolor more or I closely; sometimes even >sely as to be indistinguishable ni it. One of these plants from the middle west almost half way between the eastern and western forms of the species Dr. Rydberg has named S. perrostrata. Unfortunately, however, the difficulty in separating the two ies, though worst in the west. is not confined to that region. Some specimens from Ohio are o nearly intermediate that they ca i nined, but these are rare. There is no danger of any specimen with man leaves or in fruit being confused with any other species than S. discolor for it resembles none, but both kinds of. flowers, wl et small, resemble tho daris and the staminate are similar to those of S. Candida. Plate VIII. Salh bebbiana. Typical i fruiting; aments one with an with- a ural si ule < nlarged t hre< tin Plate IX. Salix discolor. «i7 THE WILLOWS OP OHIO. 29 Salix discolor Muhl. Pussy Willow. The pussy willow is typically a swamp shrub growing in clumps differing from those of S. sericea or S. cordata in that each plant is usually a close clump, separated from its neighbors by a distinct interval, while those species run over a considerable area in a loose clump. The stems are not ordinarily recumbent but strictly upright and straight. Twigs of swamp plants rank, metimes almost 10 mm. in thickness varying from glossy to densely tomentose, with very large well filled purple-brown buds. In less luxuriant growth the twigs may be smaller, sometimes ily. with smaller buds. Leaves varying from ovate to spatu- late, coar rate with blunt incurved teeth to entir ven htly revolute, glabrescent above, beneath glabrous and p glaucous or sometimes tomento Hair soft and wo m S. i hort, straight and ferruginous. Pussii re anthesis larger than in any o1 rid [uently this is the favorite species with the children in rch of pussies in the spring. At anthesis the staminate with their long coarse filaments are larger than any other of our wil- low pt 5. luci rpellab very large, sometimes 13 cm. in fruit, scales dark brown, capsules I -■ turn.). ite, gray pubescent to glabrate in age, pedid nearly long as the capsule but usually shorter. The flowering tim lier than any other of our willows and it uch ashorl time that it frequently happens that Salix discolor blossoms an by before oni out after it a difficulty not met with in any other of our willow.-. When the other pussy willows are found in flower it i rally in fruit so that there is little dai ■; fusing it with them. As described above Salix discolor includes forms differing from each other very strikingly. But the longer I tudy them the surer I am that, di as they are, allareone sp> The it differences arc all in characters like the shape, surface and pubescence of the li which are subject to considerable vari- ation and are to a outcome of various environ- mei >nditioi The catkii > wiry somewhat bul •m marked trees taken in flower and leaf 1 ha-, al ile to find i 1 1' 'ii 1 m ria- 1 ion ,in • !'. I X ■ iii-r lea ' t ■ 1 « >w fi broad, blunt, 1 1 1 1 5 n< mi . i. irm :i w( mid ] i :h w inter bud and nt ;i little nan I Plate X. Salix humilis and var. tristis. THE WILLOWS OF OHIO. 299 Salix eriocephala Michx. of Britton's manual includes those forms with ferruginous hair on the leaves and tonn Salix prinoides Pursh, is a narrow leaved form of the type. It is common in swamps all over the state. This s] is most difficult to separate from S. cordata in leaf but its upright habit and the coar ation as i ted with the sprawling habit and sharp-toothed leaves of S. cordata sufficient to distinguish them. As described under Salix biana it sometinn meets with that species. Narrow revo- lute-leaved forms are sometimes found which connect tl with S. humilis, probably some of them are hybrids. Salix humilis Marsh. Prairie Willow. A shrub not more than :\ meters tall with spreading often cumbent branches. Leaves oblong or spatulate, gradually nar- rowed to the base, abruptly acute and sometimes mucronate at the tij>, mostlv revolute. entire to undulate-dentate, puberulent or glabrous above, tomentose, especially on the prominently r; ed veins beneath, or glabrous and glaucous, primary veins rather distant, inclined to be horizontal, looping or oftener branching and arching, with ral costals between them, secondaries quite irregular, catkin- horn very much as those of >. discolor, long before the leaves but -mailer than in that species, from short stubby pussies, staminate 2 cm. long or shorter, carpellate asionally 4.."> em. in fruit, bracts small or none, scales dark- d above, long pilose on the back, glabrous in front, capsules elongated, often rostrate-conic in fruit, hirsute at leasl when young, pedicelled, sometimes almost 1 cm. l< ng when ripe, style distinct, red. alix humilis though common nowhere is generally distri- buted over tin' state. It will probably be fi und growing on dry ides in near) ■ »unty. In • forms Salix humilis is easily recognizable in 1. e volute leaves. In flower it is char- by tin- short stubby pus ies from which the flo\ The leaves of the ranker shool n an appearance nnilar to 1 [1 ucli bran* ' that many of tl.' tiled hybn ur herbaria came but real hybrids undoubtedly do occur. S< i the pi mnlar to V i andida, bul i rdinarily il plant while 5. Candida ha vhite wool on the und< Plati \ ■ (uppei ■■ S. numtlis; nat lltcida, > ill. ir:-< 'I Hv< Plate XL Sa-lix sericea. THE WILLOWS OF OHIO. 30] of its leaves contrasting strongly with the rich green of the upper surface. In some hybrids, however, most noteworthily Salix Candida x. S. petiolaris the resemblance is so close as to make it all but impossible to separate the two. But such hybrids usually have a distant scolloped serration derived from 5. petiolaris which is different from any form of 5. humilis. At flowering time there will be no difficulty in separating them. Salix humilis var. tristis (Ait.) Dwarf Gray Willow. A depauperate form of S. humilis with which it is ci ted by many intermediates. It may be described as smaller and hairier throughout. It is quite low (to 6 dm.) with smaller leaves (to 5 cm. long) more strongly gray tomentose, and catkins sometimes scarcely 5 mm. long. This can hardly be regarded as a distinct spi There is not a single constant character by which the two differ and what differences there are, are such would lie likely to be caused by differences in environment. h forms should be considered as varieties rather than as spec: SERICEAE. Swamp shrubs, leaves narrowly elliptic-lanceolate, com- monly tapering to both end rate with blunt cartilaginous teeth to entire, glabrou I aucous, sericeate below when young, generally blackening in drying. Aments bora before the lea-v or sh luncled, scales darkened at the tip, pili capsule led, silvery sericeate at least when young. Key. F: tm lea-. Leaves dull above, silvery silk}' beneath at least until very -'. sericea that it cannot be distinguished unless by the more ascend- ing primaries and the finer reticulations. Carpellate aments short peduncled, becoming somewhat lax in fruit, scales yellow, mostly darkened above, pilosi Jules long pedicelled, acutish, in fruit cylindric ionic, more than 4 mm. long. The i haracter of the leaves in the extreme form is very simi- lar to that of the Broad-leaved Willow with which they might be confused were it not for their narrowness. The difficulty in separating them from >'. sericea comes especially late in the sea- son when that species begins to lo pubescence and to become mor tssy on the upper surface. The catkins sometimes re- semble those of Salix bebbiana but can be distinguished by the darkened rather than yellow scales. With these exceptions the present species is not likely to be confused with any other in our a. Salix petiolaris is the rn form of Salix sericea or rather since that ■ .as later named, it is the eastern form of >'. peti ofwhichil made a variety by Andersson. It aln though that the proper treatment and that we should be doing well to return to Andersson's view but the ear; late amenl n fairly distinct. We are near the eastern border of the range of Salix petio- laris. It' rather rarely in Ohio and has been found only in the north a portion of the state, though it may extend well • ard along the lake »h< ■■ S to be expected wherever S. peti iund. Plate XII. Salix Candida and Salix petiolaris. Til!-: WILLOWS OF OHIO. ill).") Salix Candida Fluegge. Sage Willow, Hoary Willow. This little shrub seldom grows more than a meter tall. It may be recognized anywhere by its leaves which are narrowly oblong and revolute, veins deeply depressed on the upper sur- face and prominent below; under side covered with a thick wl tomentum contrasting strongly with the rich dark green of the upper surface, petiole short and stipules lacking. Flowers ap- pearing before the leaves; ovularies densely covered with silvery white wool, nearly sessile, with a very long conspicuous deep rose red style like no other of our species; staminate catkins with few bracts below them and hardly presenting diagnostic charac- ters; small and delicate resembling those of 5. bebbiana some- what but easily distinguished from them by the dark colored scales. Salix i andida is quite rare in Ohio. It was first reported by Mr. Moseley from Castalia prairie but has since been found in Wyandot county also. It ranges over the eastern and northern portions of North America. But in the west and north the leaves apparently become broader almost elliptical and not mark- edly revolute. Salix > andida hybridises with S. cordata, with 5. sericea and with S. petiolaris. At Castalia a fine series of hybrids with S. rdata and S. petiolaris may be found. CORDATAE. The cordatae are a group of shrub willows with very variable leaves, characterized by glabrous capsules borne from wooly pussies A peculiarity of the opening buds is that the inner hud le grows out beyond the outer, enveloping the base of the amenl and looking like the wing of a beetle imperfectly folded under the elytron. This so far as I know occurs in no other up and is tb an important diagnostic character at a tune when the species are particularly hard to separate. In tins te the though variable are fairly well marked but in the west the group i repre ented by a number of forms wh relationships have not rily worked out as yet. Plate XII. S i .mil. la I but ratlic-r sh( >t\ lefl natural lie and carpellate flower enlarged three times, pno graphed and bnghtem s. petiolat with si t he v;n ji in tin- ii< 1 true I, ii.iim ufe enlarged three th Plate XIII. Salix cordata. THE WILLOWS OF OHIO. 307 Key. From lea\ Leaves glossy above, heavily glaucous below, mostly broad, ascending so as to show their white under sides, glab- rous, northern. S. glaucophylla. Leaver neither glossy nor very glaucous. Leaves thick, ovate, with emarginate ' tomentose on both sides, northern. 5. adenophylla. Leaves thin, seldom pubescent above, lanceolate, gen- erally distributed. S. cordata. .vers. Bract- strongly glandular, before anthesis very villous, aments d< apsules green, shorl lied, medium sized, northern. >'. adenophylla. Bracts not glandular, tomentum quickly evanescent, pedi ■er. Br; een, obscurely serrate or entire, capsules gr< small, medium pedicelled. lata. Bracts blackening, glai it. capsules large, rostrate, becoming brown, long pedicelled, aments lax; northern. >'. glaucophylla. irdata Muhl. Heart leaved Willow. ta is the botanist's bu It is one of the most common and most variabl a it li which we have to deal, it- under no restraint of heredity and may assume ah; n and any char. 1 have seen them all the way from linear-lanceolate to orbicular, from sul>- laucous. Hut var- an- the ; ms tin .ill united Lv the flo md one cannot doubt that they all belong to a singl< Th ration of the ales alluded to above is parti" larly well marked in Salh cordata and in the rn portioi the or anywhere out of the range of the other 1 1 im S. / or 5 which it resembles much in early sprin Li triable bu1 middl alv, I usuall [n I From the former it i Plate XIII the timi Plate XIV. Salix adenophylla. THE WILLOWS OF OHIO. 309 be distinguished by its sharp fine serration contrasted with the distant blunt and often coarse teeth of the pussy willow. The habits of the two are sufheientiy dil to put aside all confu- sion when the plants arc seen together. From S. sericea it can generally be distinguished by the absence of the silvery white pub e on the under surface. The shape of the leaves is i different, in most cases in thi Les the leaf is widest near the middle; in this it is widest below the middle. Salix cordata also lacks the peculiar leaf habit i ea. The flowers come very early from small pussies. As they mature the carpellate aments come to be supported by largi es and much of the wool of the puss; E from the fruiting rachis. The anthers just before the elongati 'aments are almost as red as those of Salix sericea. The capsules are green and glabrous, the stigmas frequently red. Solid cordata angustata hnders. includes the narrow leaved forms of the species. In Ohio most plants have leaves wider than those of the typical angustata but decidedly narrower than the typical specific form. It is therefore difficult to distinguish two forms in our area and since the leaf variation may be con- sidered as accidental and without significance it is perhaps hardly advisable to separate them. Salix cordata is abundant all over the state. Its usual habitat is along ns while the other - with a similar habit and leaf are typically swamp plan; This is not to say that the presenl grows in swamps nor that S. seri and S. discolor nevi v along river banks -for they do- but that they attain their b( pment in the habitats given and are usually found To in the difficulty of dealing with Salix cordata i1 hybridises very freely. It forms with S. Candida a tin' of connecting forms. With >'. \ hybrid ir though no1 frequently as has been supposed. It is also said to mix. with >'. discolor but I have seen no unquestioned specimens from Ohio. Salix a iylla [looker. Furry Willow. A i raggling hrub o me si fata winch it resemble It looks like a xerophytic adaptation of thai i : re thicker, sh and broader, o more or less 1 . with an emarginate base, \ • or entire; ordinarily in ran th with tin- I on the twigs and 1 'late X 1 V. S fruiting aments with and w i 1 1 1> .11 1 bra< ts; a the up; natural size; Plate XV. Salix Glaucophylla. THE WILI.oW-- OF OHIO. 31 1 generally accompanied by large stipules. The flowers differ from those of our other species of cordatae in having broader bracts, much woolier before anthesis; in fruit by the denser rat- kins with shorter pedicelled capsules which are smaller than those of S. glaucophylla and rather larger than the average of 5. cordata. In typical forms the wooly broad leaves will distinguish it from everything else in our area. It is sometimes diflicull parate, however, from >'. cordata in its more tomentose forms, and intermediates seem to occur. In shape the leaves are no1 different from S. bebbiana but they will be quickly distinguished, among other things, by the sharply serrate margins of the present 5 dix adenophylla is our rarest willow. It is a plant of the Western Great Lake region and reaches its best development in Michigan, being rare east of that state. Until recently it was not supp >sed to extend into Ohio but I have seen several speci- mens collected at Erie, Pennsylvania. It should therefore ex- tend the whole length of the Ohio Lake Shore. Hut it is very scarce indeed. Though I have hunted for it on Cedar Point and m Ashtabula Countv I have seen from Ohio hut a single undoubt- ed specimen which was taken by A. I). Selby on Cedar Point. Salix glaucophylla Bebb. Broad leaved Willow. A shrub sometimes 5 m. tall, growing m clumps like >'. cor- data; leaves mostly broad, ovate or ovate-lanceolate, shiny above except in rank growth, with very tine irregular venation and stiff midribs which hold them up at an acute angle with the Stem so as to display the white under surfaces. In character the leaves mble much the narrower leaves of S. petiolaris and like them. when succulent have a strong tendency to blacken in drying. Flowers appearing before the leaves, bracts mostly narrow, glau- ■ • laminate anients generally larger than those of >. cor- data with smaller bracts; carpellate long, lax in fruit, capsules very lot m. or more) glabrous, sometimes rostrate, long ped- icelled, turning brown. Bebb labelled ome of the first material of tins species he I out, Salix cordata x 5. lucida. Tins determination though very far from correct as he himself soon recognized, is descriptive of the --i"- • [1 affinity to Salix cordata is evidenl at 01 from '.• toin "i either leaves or flowers but vet in both there is a st ron eblance to Salix lu< ida. The large thick catkins with the lot and the beautiful glossy leaves cannot but . . Plate X V s.iii x glatu ophyL I pical leaves and aments, natui d three tirrn Plate XVI. Salix purpurea and Salix pedicellaris. THE WILLOWS OF OHIO. 313 The Broad-leaved Willow is a sp< of the (ireat Lake ba- sin and in Ohio is confined to the northern part of the state. It is common along the western portion of the Lake Shore but pro- bably does not extend beyond Cleveland. Salix glaucophylla at times seems to grade into Salix cordata by what are probably a series of hybrids. In other i t is very difficult to separate from 5. discolor with which it also pro- bably hybridises. MYRTILLOIDES. BOG WILLOWS. Low shrubs with mostly elliptical, glaucous, glabrous leaves strongly reticulate veined and purplish green; aments and br. usually reddened, small, few flowered; filaments often partially united showing their affinity with the following group; a group of three or lour species all very similar to the European Salix myrtilloides. Salix pedicellaris (Anders). American Bog Willow. A low shrub seldom reaching a height of 1 meter, aereal shoots erect, slender, supported in the sphagnum by long creep- ms which run tar down into the bog putting out numerous rootlets along their length. Leaves when fully grown sometimes • l. long and 20 25mm. broad, but ordinarily smaller, elliptical, oblong, spatulate or rarely obovate, pointed at both ends, entire, slightly revolute, dark purplish-green above, slightly glaucous below, not hairy unless when very young, nearly sessile, venation picuously reticulate with meshes large and coarse consider- ing the size of the leaf. Catkins appearing with the lea- loosely flowered, no1 more than 25 mm. long in flower, bu1 sionally 5 cm. in fruit, scale short, often no longer than the nec- tary, as broad as long, filaments often more or less united; cap- sules long pedicelled, nearly always glabrous, short conic to cylin- dric-conic, obtuse, sometimes 8 mm. long in fruit, with a decided tendency to turn red or purple. Salix llaris is a northern sp< rowing in cold peat 3 win-re it may be easily recognized by its small size and ■ ! late XVI. Salix purpu S. piirpm | it .il branch with o] flowers and trim typical, the latl European specimen; natural ■ and i a] sule enlarged i hr< e t imes, phi and brightened S pedicellaris Typical h with small "\.it<- leal from the base ol iminate and carpellate il" <•! fruit ical, ' In I.im. i Hi >t ri| e . natural iminate fl< enlarged I hem, marked M . l»-t ween ' I fill pll) 314 OHIO STATE ACADEMY OF SCIENCE. culiar leaves. It it necessarily quite local in its distribution but probably grows in most of the sphagnum bogs in the state. There are specimens in the state herbarium from Williams and Portage counties and from the Licking Reservoir. I am informed by Mr. P. A. Rydberg that the typical 5. myrtilloides of Linneus was a European plant and that it differs slightly but constantly from the American species, hitherto con- sidered as the variety pedicellaris, which should consequently be considered distinct. As stated above Salix pedicellaris is easily distinguished from all the other species with which we have to deal. But more than once I have strongly suspected it of hybridising with 5. discolor which occupies the same territory. But so far I have not yet been able to satisfy myself of the hybridity and so leave the question in abeyance. PURPUREAE. Shrubs, leaves mostly oblong, oblanceolate or linear-oblan- ceolate, nearly entire, glabrescent, stamens more or less perfectly united into one with 4 anthers, capsules globose-conic, nearly sessile, silky. Salix Purpurea L. Purple Willow. A shrub reaching a height of about 3 m. putting forth a dense growth of slender wands from the larger branches. Leaves scattered or opposite, sometimes 10 cm. long and 2 cm. broad, oblong-oblanceolate, oblong or rarely elliptical, mostly broadest above the middle, abruptly acute, generally gradually narrowed to the round base, entire or obscurely serrate, glabrous, dark pur- plish-green, paler beneath, petioles short, stipules apparently absent. Catkins expanding from pussies, sessile with a few small green bracts, scales more or less pilose on the back, oblong, blunt with conspicuously purpled tips, very concave above and strongly reflexed in the staminate, less so in the carpellate, capsule broadly ovoid, silky, sessile, style very short or none. Salix purpurea is an Old World species planted in America largely for its twigs which are much used in basket work. It has been long reported in the manuals as escaped in this country but though it is fairly common over the state it is rarely that one sees a clump growing in a place where it would not likely have been planted. It certainly has not esceped to any such degree as have 5. alba and 5. fragilis. In America the carpellate plant is very rare and the species is mostly propogated by cuttings. This may account for its inability to spread as in the case of Salix babylon- ica.