Pot a NOS -Massee del. m~ Vi ^ Mier RII Wi PEN ar - 7 G.Massee LINN. OOS Journ. Bot VOI ON Fitch huh & up, ~ G. Massee del. JN.Fitch lith & imp del G Massee LINN. JOURN.—ROTANY, VOL. X¥VIT. MR. GEORGE MASSEE ON THE THELEPHOREE. 205 DESCRIPTION OF THE PLATES. Pare V. Hymenochete reniformis, Lév., nat. size. - Hymenochete tasmanica, Massee, portion of hymenium, x 400. . Hymenochete pellicula, Berk. & Broome, portion of hymenium, x 400. . Hymenochete pulcherrima, Massee, portion of hymenium and spores, x 400. . Hymenochete nigrescens, Cooke, portion of hymenium, x 400. . Hymenochete Mougeotii, Massee, portion of hymenium, x 400. . Hymenochete scabriseta, Cooke, portion of hymenium, x 400. . Hymenochete fimbriata, Ellis & Everh., spores, x 400. . . Hymenochete croceo-ferruginea, Massee. a, plant, nat. size ; 5, cystidium, X400; c, spores, x 400. Prate VI. . Corticium salicinum, Fr. a, plants, nat. size ; b, spores, x 400. - Corticium roseolum, Massee, portion of hymenium, x 400. - Corticium fatidum, Berk. & Broome. a, plant, nat. size; 4, spores, x400. . - Corticium evolvens, Fr. a, plants, nat. size; b, spores, x 400. - Corticium comedens, Fr., plant, nat. size. - Corticium calceum, Fr. a, portion of plant, nat. size; b, spores, x 400. Pare VII. . Stereum elevatum, Berk. & Cooke. a, plant, nat. size; 5, portion of hymenium, x 400. - Stereum Ravenelii, Berk. & Curt., plants, nat. size. - Stereum cyathiforme, Fr. a, plant, nat. size; b, portion of hymenium, x 400. - Stereum Tuba, Berk. & Broome, plants, X3. - Stereum radians, Fr., plants, nat. size. 206 MR. R. A. ROLFE ON THE On the Sexual Forms of Catasetum, with special reference to the Researches of Darwin and others. By R. Arren RgÉrkE, A.L.S., Assistant in the Herbarium of the Royal Gardens, Kew. [Read 21st March, 1889.] (Prate VIII.) Iw the sixth volume of the Journal of this Society, in 1862 (pp. 151-157), appeared a noteworthy paper by Darwin, bearing the title ** On the Three Remarkable Sexual Forms of Catasetum tridentatum, an Orchid, in the possession of the Linnean Society.”* The purport of this paper was to show that the species in question had been observed to produce three distinct kinds of flowers, belonging to the same number of supposed genera, all on the same plant. Each of the three forms was represented by a woodcut. The explanation given of this curious phenomenon was that the three forms represented, respectively, the male, female, and hermaphrodite states of the same species. This paper was a remarkable one in many respects. In the first place, it established conclusively the fact—already more than sus- pected by Schomburgk—that the so-called sportive character of Catasetum, or the curious habit of its species of suddenly pro- ducing flowers of a totally different kind (usually termed * monsters ") on the same plant, was simply an abnormal combi- nation of different sexual forms in the same individual. In the second place, it described, fully aud graphically, the remarkable structural adaptations by which self-fertilization was prevented and cross-fertilization (by insect agency) secured. And, in the third place, it indicated a condition of things which, I venture to think, is unparallelled in the entire Vegetable Kingdom, namely —the production, by one and the same species, of one kind of flowers, wholly male, of a second kind, wholly female, dependent on a highly complex process for their fertilization ; and of yet a third kind, hermaphrodite in character, yet not cleistogamous—in other words, still highly specialized and incapable of self- fertilization. * The accounts given by the same author in his work ‘ On the Various Con- trivances by which British and Foreign Orchids are Fertilized by Insects,’ ed. 1, pp. 236-247, and in Ann. Sc. Nat. sér. 4, xix. pp. 247-255, t. 12. fig. A, are substantially identical. m E. SEXUAL FORMS OF CATASETUM. 207 This third point, however, was not established on the same firm foundation as the two former ones; and to this alone may be attributed the appearance of the present paper. The whole subject is one of peculiar fascination and clothed with a balo of romance; but, as soon as I thought seriously on tbe matter, the altogether anomalous character of the so-called hermaphrodite flowers was apparent; and I was so impressed with the impro- bability of Darwin's explanation being the correct one, that I determined, as soon as the opportunity presented itself, of going over the whole ground again. The results have been of a somewhat, though perhaps not altogether, unlooked for nature. - Briefly, they may be stated, at the outset, as follows :— l. Only two kinds of flowers, male and female, are produced by the species of Cafasetum investigated by Darwin; and with the exception of the section Pseudocatasetum (vide infrà, p. 224), and a doubtful example noted on pp. 214, 215 (C. Gnomus), this appears to be the normal habit of the species generally. 2. The so-called hermaphrodite form is simply a male—the corresponding sex of the female investigated by Darwin ; and the two do not belong to Catasetum tridentatum at all, but to C. barbatum, Lindl. 3. Darwin's male form is the only one which belongs to Cata- setum tridentatum,—that is, limiting the use of the term to the forms actually examined by him (and figured); for the female of this very plant had been previously figured *, and the figure was doubtless known to the author, though supposed to be identical with the one he himself had examined. From this it will appear that the author has fallen into some error, the nature and origin of which must now be traced. The statement that three distinct kinds of flowers had been observed all growing on the same plant f is really based on the observations of Schomburgk; but these were unfortunately mis- understood by Darwin. What Schomburgk really stated was, that three kinds of flowers were produced by (apparently) one * Lindley, Bot. Reg. t. 1752. . + “Botanists were astonished when Sir R. Schomburgk [Trans. Linn. Boc. xvii. p. 551] stated that he had seen three distinct forms, believed to constitute three distinct genera, namely, Catasetum tridentatum, Monachanthus viridis, and Myanthus barbatus, all growing on the same plant."— Darwin, Journ. Linn. Soc, vi. p. 151. Q2 208 MR. R. A. ROLFE ON THE and the same species *, not by the same plant (individual). The difference may at first sight seem an unimportant one; but a moments consideration will show that this is not the case. Had the three kinds of flowers been observed on the same individual f, no amount of reasoning could have explained away the fact. As it is, however, the possibility of two distinct species having been accidentally confused together presents itself; and this is what has really taken place. It may appear singular that two such very different looking plants as Cutasetum tridentatum and C. barbatum—that is the male forms, to which the names were exclusively applied,—should ever have been confounded together ; but on looking at the females only of the two species, there appears nothing very remarkable in the fact that they should have been thought identical t. Here we have the key of the situation. The female forms of the two—and indeed of several others—are remarkably alike; consequently the fact that they belonged to quite distinct species was entirely overlooked, not only by Darwin, but by other writers. * “In a letter which I had the pleasure to address to Mr. Bentham, on the 28th of June last year, I informed him of a remarkable Orchidaceous plant, from appearance a Monachanthus, which on one side of the bulb produced a scape with six flowers of Monachanthus viridis and two of Myanthus barbatus, while a second scape on the same bulb had twenty-five blossoms of the Myan- thus barbatus. .. . . The thought impresses itself, therefore, forcibly upon me, that the genera Monachanthus, Myanthus, and Catasetum form but one genus; and in this conclusion I am borne out by the following observations. “A vigorous plant, which produced at its former state of inflorescence the flowers of Monachanthus viridis, had two months ago a scape with flowers of Catasetum tridentatum. . ... Mr. Bach, an enthusiastic collector of Orchida- ceous plants, sowed the seed of Monachanthus viridis on a decayed trunk of an Erythrina. Among these plante, one produced a scape with the flowers of Catasetum tridentatum..... The evidence of the present plant, which has caused these remarks, would likewise include the genus Myanthus in the group.” —Schomburgk, in Trans. Linn. Soc. xvii. p. 551. t It is clear that when Darwin said “plant,” he did not mean “ species ;” for after describing Catasetum tridentatum and Monachanthus viridis, he pro- ceeds to say : “ We now come to the third form, Myanthus barbatus, often borne on the same plant with the two preceding forms.’—Darwin, Journ. Linn. Soc. vi. p. 156. 1 For example, Catasetum cristatum, var. monstrosum, Hook. Icon. Plant. ii. t. 177, is a plant sent to the author, in alcohol, by Dr. Nimmo. The nine uppermost flowern are female, the three lower ones male. The author remarks that he is not sure if Rot. Reg. t. 1752 [really the female of C. tridentatum) is not the same as the cnc now figured, with all its flowers metainorphosed. SEXUAL FORMS OF CATASETUM. 209 That Darwin relied very largely on the statements of Schom- burgk and other authors is evident; and this leads us to examine their writings in detail. In the * Botanical Register ' for 1836 (t. 1752) a so-called Mona- chanthus viridis had been figured by Lindley, from a garden specimen; and with this Schomburgk rightly identified a form he had actually seen produced by a plant of Catasetum triden- fatum*, But he also wrongly identified with the same a similar form, produced by a plant of Myanthus barbatus*. Thus, both being thought to be Monachanthus viridis, and having been produced by plants hitherto considered distinct, all the forms were thought to belong to one and the same species. We must now turn to the origin of Monachanthus viridis. In the ‘ Botanical Register’ for 1832, in a note under t. 1538, Lindley speaks of two new genera in the collection of Sir William Hooker; which he then describes. One of these is Monachanthus, which is said to differ from Catasetum in having the column altogether without cirrhi. The single species, Mona- chanthus viridis, he describes as growing on trees in the Corcovado, near Rio de Janeiro. This very specimen is preserved at Kew. It consists of a water-colour drawing of the entire plant, together with a single dried flower. Who sent it, I have failed to dis- cover; but it is labelled, * From the Corcovado, near Rio de Janeiro; grows on trees. The bulbs and leaves are similar to those of a different flower.” It is labelled by Lindley himself, " Monachanthus viridis, Bot. Reg. 1538." The statement that the bulb and leaves are similar to thoee of a different flower, leads us to examine the other genus described by Lindley ; for to it the note applies. The genus in question is Myanthus, which is said to be closely allied to Cafaeetum, but with the cirrhi situated at the base of the column. The single species, Myanthus cernuus, is said to have been found or trees in the neighbourhood of Rio de Janeiro. Lindley then adds :—" Perhaps Catasetum cristatum would be better referred to this genus.” This specimen is also preserved at Kew. It consists of a com- plete water-colour drawin g, witha dried raceme in addition ; four flowers out of about fourteen alone remaining. It is labelled, “From the neighbourhood of Rio de Janeiro; grows on trees. Lindley’s ticket says, “Myanthus cernuus, B. Reg. 1538.” The * See footnote on preceding page. 210 MR. R. A. ROLFE ON THE two drawings were evidently made by the same hand and at the same time, for the colours used as well as the workmanship are so identical as to leave no doubt on the subject. Looking at the two, there cannot be the slightest doubt that they repre- sent respectively the male and female sexes of the same species. We now return to Lindley's published figure of so-called Monachanthus viridis*, which appeared four years after the original description was published; but only to find that a second species is represented under the same name. Although Lindley here states this to be the original species on which the genus was founded, it is clear that such was not the case. It is simply a case of mistaken identity. At this time the connection between the plant figured and Catasetum tridentatum, though pointed out to Lindley, was not admitted by him, as he afterwards states t. This appears to me to have introduced the first element of confusion into the question. The original Monachanthus viridis, Lindl., is evidently the female of Myanthus cernuus, Lindl., as above pointed out; but now Lindley complicates matters by figuring the female of a second species; namely Catasetum tri- dentatum, as Monachanthus viridis. The question becomes still further involved when Schomburgk figures the female of yet a third species, namely Myanthus barbatus, also as Monachanthus viridis. Nor is this all; for in 1844 Lindley observed of Catasetum cristatum, * Has been found to sport into C. tridentatum" T1. In fact, the idea appears to have been fast gaining ground that the ordinary conceptions about species could not be applied to Catasetum, which might sport into almost anything; for when * ‘This is the original species on which the genus was founded. Jts habit is so Die Catasctum tridentatum, that we long doubted whether it ought to be generically separated. .... Our figure was made from a specimen communi- cated last November from Wentworth, by permission of Lord Fitzwilliam. We are not aware at what time, or by whom it was imported, but it has probably been taken for a green form of the common Ca/asetum tridentatum, and conse- quently no record has been kept of it."— Lind/ey, Bot. Reg. t. 1752. t Even when Lord Fitzwilliam assured me that it was beyond all doubt an accidental sport of Catasetum tridentatum, I still adhered to my idea that an imported plant of Monachanthus viridis had been accidentally taken for the latter common piant." — Lindley, Bot. Reg. t. 1951*. 1 Lindley, Bot. Reg. xxx. Misc. p. 39. SEXUAL FORMS OF CATASETUM. 211 Sir William Hooker figured Catasetum Naso, Lindl., he speaks of giving the name furnished by Lindley, with a feeling that it might be a sport of the original C. tridentatum*. It may be observed here that the idea of three supposed genera on the same plant did not actually originate with Darwin; for Lindley has a similar statement with respect to another species, namely C. cristatum. Speaking of an abnormal specimen of that plant, he refers to it as combining in its own proper person no fewer than three supposed genera, Myanthus, Monachanthus, and Catasetum t. The statement, however, is singularly unfor- tunate and misleading ; for on referring to the accompanying plate only two kinds of flowers are represented, namely, nine females on the upper part of the raceme, and seven males underneath. It is to this specimen that Lindley alludes when he speaks of Catasetum cristatum sporting into C. tridentatum, though this would appear to have been an after-thought ; for at the time he only speaks of it as changing into a Monachanthus, allied to M. viridis. The singular phenomenon just described by Lindley was not now observed by him for the first time. As long previously as 1826, when he figured this very species, he observes :—“ The un- importance of the peculiarity which exists in the labellum is manifested in a singular manner by a curious monster of this plant, whieh we have observed in an individual in the Horti- cultural Society’s Garden. Among flowers of the ordinary kind two or three others were observed in which the labellum was precisely of the same nature as that of Catasetum tridentatum; that is to say, destitute of the crested appendage, and perfectly galeate and naked f." It is unfortunate that these were not represented on the plate. This is interesting as being the first recorded evidence of the production of two kinds of flowers by the genus; and had Lindley * Hooker, Bot. Mag. t. 4792. . + “In November 1836 His Grace the Duke of Devonshire was so kind as to put into my hands the extraordinary flower represented in the accompanying plate, which may be regarded as one of the greatest curiosities that our gardens ever produced... .. It is that of a plant of Myanthus cristatus changing into a Monachanthus related to Monachanthus viridis, and combining in its own proper person no fewer than three supposed genera, Myanthus, Monachanthus, and Catasetum.” — Lindley, Bot. Reg. t. 1947 A (text numbered ** 1951* "). 1 Lindley, Bot. Reg. t. 966. 212 "MR. R. A. ROLFE ON THE grasped its full significance, it is certain that he would not afterwards have described Monachanthus and Myanthus at all. As it was, however, some years afterwards he tells us he could not believe his own words, and concluded some mistake had been made*. Even when finally abandoning the two supposed genera, he had not the slightest idea of the significance of the facts ob- served t. Nor does he ever appear to have arrived at a solution of the difficulty, as the following will show :—‘ There is a circum- stance observed by Mr. Schomburgk ... which is very curious, and deserves to be recorded. Ina letter... he says... ‘Are you aware that Catasetum and Myanthus are not seed-bearing, but that Monachanthus bears seed abundantly ?' Ido not know what conclusion to draw from this statement; but it would be a curious fact if, as Mr. Schomburgk’s observation would seem to imply, the species of Catasetum and Myanthus should prove to be sterile states of Monachanthus " 1. It is not a little remarkable that Schomburgk, coming so near, as he did, to guessing the truth, should not have seen that he had two females confused under the name Monachanthus viridis ; but the idea of the plant having two quite different kinds of males does not appear to have aroused his suspicions. The following note will show what his ideas on the subject were :— * Here we have traces of sexual difference in Orchidaceous flowers. I have seen hundreds of Catasetum tridentatum on savannahs adjacent to the lake Capoeya (Arabisce coast of Essequebo), without ever finding one specimen with seeds, while those bulbs which, according to Dr. Lindley's description, belonged to Monachanthus viridis, astonished me by their gigantic seed-vessels ” $. 7 * « This, I repeat, appeared to me so extraordinary a statement, especially as after seven years it had never been corroborated by any other case of the same kind, that I concluded I must have made some mistake.” —Lindley, Bot. Reg. t. 1947 A. t “The necessary consequence of this is that the supposed genera Myanthus and Monachanthus must be restored to Catasetum ; and I have no doubt now, although no proof has been seen of it, that Mormodes must share the same fate. But which of the species have their masks on, and which show their real faces, I certainly will not at present presume to guess.”—Lindley, Bot. Reg. t. 1947 A. 1 Lindley, Bot. Reg. xxiv. t. 31. $ Schomb. in Trans. Linn. Soc. xvii. p. 552. SEXUAL FORMS OF CATASETUM. 213 It remained for Darwin to investigate the structure of the flowers, and to show by what remarkable mechanisms the work of fertilization was accomplished. In the case of the male and female flowers, the conclusions arrived at are thoroughly satis- factory; but the same cannot be said respecting the so-called hermaphrodite form. Of this he observes, after describing the floral envelopes :—“ The antennæ are not so long as in the male C. tridentatum, and they project symmetrically on each side of the horn-like projection at the base of the labellum, with their tips (which are not roughened with papillæ as in the male flower) almost entering the medial cavity. The stigmatic chamber is of nearly intermediate size between that of the male and female forms; it is lined with utriculi charged with brown matter. The straight and well-furrowed ovarium is nearly twice as long as in Monachanthus, but is not so thick where it joins the flower ; the ovules are not so numerous as in the female form, but are opaque and pulpy after having been kept in spirits, and resemble them in all respects. I believe, but dare not speak positively as in the case of the Monachanthus, that I saw the nucleus projecting from the testa. The pollinia are about a quarter of the size of those of the male Catasetum, but have a perfectly well-developed dise and pedicel. The pollen-masses were lost in the specimens examined by me; but fortunately M. Reiss has given, in the ‘Linnean Transactions,’ a drawing of them, showing that they are of due proportional size, and have the proper folded or cleft structure; so that there can hardly be a doubt that they are functionally perfect"*. These results were obtained from a specimen preserved in spirit, but during last autumn Mr. F. W. Moore, Curator of the Glasnevin Botanic Garden, was good enough to send me a living raceme of the same identical form. After careful examination I can only say that the whole series of female organs are as imper- fect as in any other male Catasetum, while, on the other hand, the male organs themselves differ in no respect whatever from those of several allied species. In other words, that the form in question is a male only, and belonging to the group in which Darwin showed that both the antenne are sensitive ; as opposed to the other group (including C. tridentatum), in which one only is sensitive, the other being turned round near the base of the * Darwin, in Journ. Linn. Soc. vi. p. 156. 214 MR. R. A. ROLFE ON THE column and functionless. This being the case, Darwin's other remarks upon Myanthus* naturally fall to the ground. Hitherto the instances of so-called "sporting" in Catasetum have not presented any insuperable diffieulty on careful investi- gation, but have proved to be the occasional abnormal combination of both sexes in the same individual; whereas the species are normally dicecious,—so much so that in the majority of them the second sex is at present unknown. But now we come to another recorded instance of the production of three kinds of flowers on the same individual, of which I at present fail to discovera satis- factory explanation. Speaking of a supposed new species, C. heteranthum,— which, by the way, has proved identical with C. Gnomus, Linden & Reichb. f.,—Rodrigues remarks that he has observed three kinds of flowers borne on the same individual, corresponding to the three “subgenera ” Catasetum, Myanthus, and JMonachanthus t. Each of the three is pretty fully de- scribed, but, strange to say, there is not a word about sexual dif- ferences, neither in this case nor in that of other species of which two kinds of flowers were observed. In fact, the author does not * “As we thus see that both the male and female organs are apparently perfect, Myanthus barbatus may be considered as the hermaphrodite form of the same species, of which the Catasetum is the male and Monachanthus the female. * It is not a little singular that the hermaphrodite Myanthus should resemble in its whole structure much more closely the male forms of two distinct species (namely, C. saccatum and, more especially, C. callosum) than either its own male or female forms."— Darwin, in Journ. Linn. Soc. vi. p. 156. ** In Catasetum we have three sexual forms, generally borne on separate plante, but sometimes mingled together; and these three forms are wonderfully dif- ferent from each other—much more different than, for instance, a peacock is from a peahen.”— Darwin, l. c. p. 157. t “In eodem individuo profert hiec species, in eodem caule seu caulibus aliis, generis tres subdivisiones.”—Rodrigues, Gen. et Sp. Orch. Nov. i. (1877), p. 127. * Sur une Ayurana (Salix Humboldtiana, Mart.) dans Amazonas, j'ai rencontré un exemplaire présentant un fait d'hétéranthie trós-curieux; du même pseudo- bulbe sortaient trois tiges florales dont chacune portait séparément des fleurs des sous-genres Catasetum, Myanthus, et Monachanthus. Les fleurs, outre la forme, avaient le coloris trés-différent. Plus tard, j'ai rencontré dans Vigapé du Yanauary, un autre individu portant sur la méme tige le Catasète et le Myan- thus." — Rodr. l. c. p. 128. "J'ai vu dans .. . . plusieurs endroits de l'4mazonas, le méme individu, émettre non svuleument le Myanthus et le Catasetum en racemes différents, mais aussi le même racéme offrir le Monachanthus avec les deux autres formes.” — Rodr. l. c. p. 206. SEXUAL FORMS OF CATASETUM. 215 appear to regard any of them as sexual forms, for he constantly speaks of them as representing “ sections " or “ subgenera.” Nor is there a word as to those most essential points—the presence or absence of the pollinia, and of the antennz of the column, both of which strengthen my suspicions on the point just men- tioned. J udging, however, by the description, the “ § Monachan- thus” represents simply the female form, and thus presents no special difficulty. The “§ Myanthus” likewise appears to repre- sent the male form, and for the following reasons:—In the first place it agrees with the plant cultivated in European gardens, and of which the Messrs. Veitch, of Chelsea, were so good as to send me a complete raceme and leaf only a very few days ago. That it is this plant Rodrigues himself asserts, in a later note (J. c. p. 206), for he points out that it is the one figured by Reichenbach (Xen. Orch. ii. p. 171, t. 170. fig. 5), also by André (Ill. Hort. xxiv. n. s., t. 170), though he lays claim to his own later name superseding the original one, on the ground that he first described the other forms of the species, and that it it known under the name in Brazil. Here, then, we have the Myanthus form, and consequently can examine its characters. First, it is not a Myanthus at all, confining the term to the group to which it was originally applied, and which may be adopted (in à somewhat altered sense) for a large and very natural section of the genus. It belongs to another group, hereafter defined as Eucatasetum, in which the lip of both sexes is superior. The ovary is atrophied, as in other males of the genus, the column eompa- ratively elongated, the pollinia perfect, one of the antennz stand- ing forward and highly sensitive, and the other turned round near the base of the column and functionless, as in others of the section. Having disposed of the male and female forms of the species, there remains the so-called * § Catasetum” form, which to me is a complete mystery. The complete description is here appended *, and it will be observed that it is just the essential * “8 Catasetum, scapo erecto, crasso, 3-5-floribus ; perianthio explanato ; sepalis oblongis, acutis, concavis, patentissimis; petalis oblongis, ad basin atten- uatis, cum marginibus recurvis; labello obconico, cam marginibus crispis, dentatis, recurvis, acuto in fronte; gynostemio brevi, mucronato. “ Les sépales sont vertes, fortement fasciées de brun-pourpre, les pétales vert mouchetés aussi de brun-pourpre, le labelle jaunâtre trés-finement moucheté de la méme couleur, ainsi que le gynostéme, qui est aussi vert. La tige est verte." —Rodr, l. c. p. 127. 216 MR. R. A. ROLFE ON THE points, which alone might enable one to form some conclusion, that are omitted. Under “§ Myanthus," however, we read, “ la- bello conforme Catesetum cum parte obconica tenuiore; gyno- stemio elongato mucronatissimo," so that the most obvious dif- ferences pointed out of this “§ Catasetum " form are the shorter few-flowered raceme, the more spreading segments, the broader conical portion of the lip, the shorter column, and somewhat different coloration. What the biological significance of these differences may be I cannot conceive, and I can only hope that some Brazilian naturalist, into whose hands this paper may fall, will be good enough to forward to Kew such material as may enable the question to be decided. A dried specimen showing the three kinds of inflorescence described would be invaluable, or even the three kinds if borne on separate plants, though in the latter case there is, perhaps, some danger of two species being confused, as on former occasions. I propose to conclude this paper, firstly, by enumerating all the species of Catasetum of which both sexes are known, giving references to descriptions and figures of each sex ; and, secondly, by outlining the principal subdivisions of the genus, indieating the more prominent characters and best known species of each. Many species are yet very imperfectly known, and I should be extremely obliged to any one who will try to remedy this by forwarding to Kew specimens, especially of any instances of the produetion of the second sexual form of a species. I may here point out that in the previous part of this paper the names used by the several authors in question are adopted, but in the following enumeration the earliest unappropriated name of each species is invariably given precedence. Thus Cafa- setum tridentatum, Hook., has proved to be the original species upon which the genus was founded, namely C. macrocarpum, Rich. the species of Myanthus take the same specific name under Cata- setum; while Monachanthus viridis is suppressed altogether, as the name applies to the females of three distinct species of the genus. ENUMERATION OF SPECIES OF CATASETUM OF WHICH BOTH SEXES ARE KNOWN. 1. CaTASETUM ATRATUM, Lindl. d . C. atratum, Lindl. in Bot. Reg. xxiv.(1838), Misc. p. 61; id. t. 63; Bot. Mag. t. 5202. pec mem. — n — 0 SEXUAL FORMS OF CATASETUM. 217 9. C. atratum, Lindl.; Rodr. Gen. et Sp. Orch. Nov. ii. (1881), p. 220 (§ Monachanthus). 2. CATASETUM BARBATUM, Lindl. d. Myanthus barbatus, Lindl. Bot. Reg. xxi. (1836) t. 1778; Bot. Mag. t. 3514; Past. Mag. Bot. ii. p. 124, with plate; Kn. § Weste. Fl. Cab. i. t. 87; Schomb. in Trans. Linn. Soc. xvii. p. 551, t. 29 (also 9); Darwin in Journ. Linn. Soc. vi. p. 153, fig. 28; id. Fertiliz. Orch. ed. 1, p. 239, fig. 28 B; id. in Ann. Sc. Nat. sér. 4, xix. p. 247, t. 12. | fig. A 4.—Catasetum barbatum, Lindl. Bot. Reg. xxx. | (1844), Misc. p. 38; Reichb. f. in Walp. Ann. Bot. vi. p. 570. - 9. Monachanthus viridis, Schomb. in Trans. Linn. Soc. xvii. | (1837), p. 551 (in part), t. 29 (also d); Darwin in Journ. Linn. Soc. vi. p. 153, fig. 24; id. Fertiliz. Orch. ed. 1, | p. 239, fig. 284; id. in Ann. Sc. Nat. sér. 4, xix. p. 247, | t. 12. fig. A 3. Very closely allied to C. cristatum. Schomburgk confounded | the female of this species with that of C. tridentatum, referring | all to Monachanthus viridis, Lindl. The female (so-called Mona- | chanthus viridis) investigated by Darwin also belongs to the | present species. 3. CATASETUM CALLOSUM, Lindl. d. C. callosum, Lindl. Bot. Reg. xxvi. (1840), Mise. p. 77; id. xxvii. t. 5; Bot. Mag. t. 4219 (var. grandiflorum, Hook.) ; id. t. 6648. 9. C. callosum, Lindl.; Reichb. f. in Walp. Ann. Bot. vi. (1861), p. 577. 4. CATASETUM CERNUUM, Reichb. f. d. Myanthus cernuus, Lindl. Bot. Reg. xviii. (1832), sub t. 1538; id. t. 1721; id. Gen. § Sp. Orch. p. 155.— Catasetum trifidum, Hook. Bot. Mag. 1x. (1833), t. 3262. —C. cernuum, Reichb. f., in Walp. Ann. Bot. vi. (1860), p. 570; Bot. Mag. t. 5399. 9 . Monachanthus viridis, Lindl. Bot. Reg. xviii. (1832), sub t. 1538 ; id. Gen. & Sp. Orch. p. 157 (not elsewhere).— Catasetum cernuum, Reichb. f. ; Rodr. Gen. et Sp. Orch. Nov. ii. p. 218 (§ Monachanthus). 218 MR. R. A. ROLFE ON THE The female of this species is clearly the original Monachanthus viridis, as pointed out at p.210. Lindley himself afterwards confounded the female of C. tridentatum with it, while Schom- burgk confounded together both the latter and the female of C. arbatum, under the same name. I have only seen the original drawing and a single dried flower, but the so-called “ § Mona- chanthus " form of Catasetum cernuum, cited above, is clearly the same. 5. CATASETUM CRISTATUM, Lindl. g. C. cristatum, Lindl. Bot. Reg. x. (1824), sub t. 840; id. t. 966 ; Var. monstrosum, Hook. Ic. Pl. ii. t. 177 (also 9 ). W, —Myanthus cristatís, Lindl. Bot. Reg. xxiii. (1887), t. 1947 a, text 1951* (also 9 ). 9. Monachanthus cristatus, Lindl. Bot. Reg. xxiii. (1837), t. 1947 a, text 1951* (also ¢ ).—Catasetum cristatum, var. monstrosum, Hook. Icon. Pl. ii. t. 177 (also d ). This species is interesting, as the one in which the occurrence of a second kind of flower in the genus Catasetum was first ob- served, though the significance of the phenomenon was until long afterwards quite unknown. At t. 966 of the ‘Botanical Register ` (in 1826) Lindley speaks of “a curious monster" which bore, among flowers of the ordinary kind, two or three others in which the labellum was of the same nature as that of C. tridentatum. 6. CarAgETUM DanwiNIANUM, Rolfe. (Plate VIII.) d. C. Darwinianum, Rolfe, in Gard. Chron. ser. 3, v. p. 391. — C. fuliginosum, Rolfe, in Gard. Chron. ser. 8,iv. (1887) p. 473 (non Lindl.). 9. C. Darwinianum, Rolfe, in Gard. Chron. ser. 9, v. p. 394. —C. fuliginosum, Rolfe, in Gard. Chron. ser. 3, iv. (1887), p. 473 (non Lindl.). This is a plant which flowered in the Kew Collection during the autumn of 1888. At the time I identified it with C. fuligi- nosum, Lindl., described from a female specimen only, of which the native country is unknown. But on working up the present paper and comparing the dried specimens side by side, I think the identifieation erroneous, and have therefore proposed the new name above cited. The plant, which is a native of British Guiana, is represented in the accompanying Plate, for which I am indebted to the Bentham Trustees. a ——— RN SEXUAL FORMS OF CATASETUM. 219 7. CATASETUM DELTOIDEUM, Lindl. d. Myanthus deltoideus, Lindl. Bot. Reg. xxii.(1836),t. 1896; Bot. Mag. t. 3923. 9. Catasetum deltoideum, Lindl. Bot. Reg. xxvi. (1840), Mise. p. 71. The female of this has not been figured, but there is an excellent specimen in Lindley's Herbarium. 8. CatasetuM GNowvs, Linden § Reichb. f. d. C. Gnomus, Linden § Reichb. f. ex Reichb.f. Xen. Orch. ii. (1873), p. 171, t. 170. fig. 5 ; IIL. Hort. xxiv. n. s. t. 270. —C. heteranthum, Rodr. Gen. et Sp. Orch. Nov.i. (1877), p. 127 (§ Myanthus ; also ? § Cutasetum), et p. 205. F. C. heteranthum, Rodr. Gen et Sp. Orch. Nov. i. (1877), p. 128 (8 Monachanthus), et p. 205. I have only seen the male of this species, the so-called “ $ My- anthus " form of Rodrigues. Three distinct forms are pointed out by that author, on which point see remarks, supra, pp. 214, 215. 9. CATASETUM MACROCARPUM, Rich. d . C. macrocarpum, Rich. er Kunth, Syn. Pl.i.(1822) p. 331, in footnote; Kunth, in Humb. & Bonpl. N. Gen. et Sp. vii. p. 158, t. 231 (also fruit, with restored female flower) ; Warn. Z Will. Orch. Alb. t. 189; Ill. Hort. n. s. t. 619. — C. tridentatum, Hook. Exot. Fl. ii. (1825), tt. 90, 91; Bot. Mag. tt. 2559, 3329 ; Lindl. Bot. Reg. t. 840 ; Rchb. Fl. Exot. i. t. 60; Schomb. in Trans. Linn. Soc. xvii. p. 551 (in part); Darwin, in Journ. Linn. Soc. vi. p. 192, fig. 1; id. Fertiliz. Orch. ed. 1, p. 232, fig. 27 ; id. in Ann. Sc. Nat. sér. 4, xix. p. 247, t. 12. figs. A 1, 2.—C. Cla- veringii, Lindl. Bot. Reg. x. (1824), t. 840; Lodd. Bot. Cab. t. 1344 ; Reichb. Fl. Exot. ii. t. 122.—C. floribundum, Hook. Exot. Fl. ii. (1825), t. 151. 9. Monachanthus viridis, Lindl. Bot. Reg. xxi. (1936), t. 1752 (not elsewhere).—Catasetum macrocarpum, Kunth, / < in Humb. § Bonpl. N. Gen. et Sp. vii. p. 158, t. £31 (also 3). The male plant is very variable, and has received several names. The female is not the original M. viridis, though it has been con- founded with it and with the females of other species. It is 220 MR. R. A. ROLFE ON THE remarkable that on Kunth's plate, above cited, there is a figure of a female flower attached to the fruit, but, being restored from the withered flower, it is inaccurate in some respects, notably the column and lip, which have clearly been helped out from the male flower. In Lindley's Herbarium is a most interesting specimen, with eight female flowers below and two males at the apex of the raceme. 10. Carasrrum Naso, Lindl. d. C. Naso, Lindl. Bot. Reg. xxix. (1843), Misc. p. 71; id. xxx., Mise. p. 86 ; Bot. Mag. t. 4792. 9. Thereis a MS. drawing at Kew of a plant which flowered at Syon House, which shows both sexes of this species. The drawing is unfinished, but a single detached flower of either sex is finished in colours. The male has the petals somewhat resem- bling the lip in character, though otherwise normal. Under- neath is written, “ Another apparent var. of Cat. Naso—but with remarkable petals.” The female is quite normal, and under it is written “ Monachanthus viridis." This very drawing is the original of Bot. Mag. t. 4792, but for some reason the female flower is altogether omitted from the plate. Sir William Hooker remarks :—“ Drawings of the two kinds of flowers here given . were submitted to Dr. Lindley, who considers them to be varieties of C. Naso; as such we represent them, with a feeling, however, upon our minds that they may all be sports of the original C. tridentatum.”. I may add that in the published plate all the flowers represented on the raceme are males in the normal condition, and that the detached flower represents a peloriate state of the same, with fimbriate petals, as pointed out above. 11. CaraskTUM ŒRSTEDII, Reichb. f. d. C. Erstedii, Beicht, f. in Bonpl. iii. (1855), p. 218 ; id. in Walp. Ann. Bot. vi. p. 565. 9. C. (Erstedii, Reichb. f.; Walp. Ann. Bot. vi. (1861), p. 577. 12. CaTAsETUM PILEATUM, Reichb. f. d. C. pileatum, Reichb. f. in Gard. Chron. n.s. xvii. p. 492 ; id. xxvi. p. 616 ; Sander, Reichenbachia, ii. p. 91 (also gl t. 90.—C. Bungerothii, N. E. Br. in Lindenia, ii. (1886), SEXUAL FORMS OF CATASETUM. 221 D 21, t. 57; id. tt. 104 (var. Pottsianum), 116 (var. aureum); Bot. Mag. i. 6998; Warn. 4 Will. Orch. Alb. t. 352; Ill. Hort. xxxiv. t. 10; Gard. Chron. ser. 3, i. p. 189, with plate, p. 142, fig. 32; N. E. Br. inGard. Chron. ser. 3, v. p. 461, fig. 83 a. 2. C. pileatum, Reichb. f. ; Sander, Reichenbachia, ii. p. 91 (also d).—C. Bungerothii, N. E. Br. in Gard. Chron. ser. 3, v. p. 461, fig. 832; Rolfe, in Gard. Chron. ser. 3, vi. p. 466. A single female flower was produced on the same raceme with those of the other sex during last autumn, in the collection of the Société de l’ Horticulture Internationale, at Brussels, and was sent to Kew by M. Lucien Linden, where it is now preserved. It is very distinct in appearance, and could not well be mistaken for any other at present known. 13. Caraserum Renew, Rodr. d. C. Regnellii, Rodr. Gen. et Sp. Orch. Nov. ii. (1881), P- 219 (S Myanthus). F. C. Regnellii, Rodr. l. c. p. 219 (§ Monachanthus). Only known to me by the description, from which it is clear that the two forms described are simply the sexual states of the species. 14. CarASETUM SANGUINEUM, Lindl. d. C. sanguineum, Lindl. in Part, Fl. Gard. ii. (1851-2), P. 168, fig. 225; Pescatorea, i. t. 16.—Myanthus san- guineus, Linden, ex Lindl. in Past. Fl. Gard. ii. p. 168 (in note). 9. C. sanguineum, Lindl. ; Planch. Hort. Donat. (1894-8), p. 136 ; Heichb. f. in Gard. Chron. n. s. ix. p. 104. The female of this species I have not seen, but it is described in the places above cited. 15. Carasrrum UMBROSUM, Rodr. d. C. umbrosum, Rodr. Gen. et Sp. Orch. Nov. i. (1877), P- 129 (§ Catasetum). 9 . C. umbrosum, Rodr. l. c. p. 129 ($ Monachanthus). Only kuown to me by the description, from which it is clear the two forms described are the sexual statesof the species. The LINN. JOURN.— BOTANY, VOL. XXVII. R 222 MR. R. A. ROLFE ON THE position of the lip is not mentioned, but from the characters given I believe it belongs to the section Myanthus. ) 16. CaTASETUM VARIABILE, Rodr. d. C. variabile, Rodr. Gen. et Sp. Orch. Nov. ii. (1881), p- 217 (S Myanthus). 9. C. variabile, Rodr. l. c. p. 217 (§ Monachanthus). Only known to me by the description, from which it is clear that the two forms described are simply the sexual states of the species. SUBDIVISIONS op THE GENUS CATASETUM. No attempt has hitherto been made to separate the species of Catasetum into natural groups; though, with the clearing away of the mystery which has surrounded them and the accumulation of a more complete series of specimens, it is certain that such a subdivision will be not only possible, but absolutely necessary, in order to get something like a clear idea of their relations to each other. There appears to me to be four very distinct sections under which the species may be grouped ; and I have not seen a single species of which there can be any doubt as to its position. The second group, however, Myanthus, absorbs the majority of the species. The following list only includes some of the better known species, and is not intended to be exhaustive. Section I. EucarAsETUM.—Diccious; lip superior in both sexes, generally more or less galeate in the male, always so (as far as known) in the female; rostellum prolonged below into a pair of slender cirrhi, the antennæ, one of which is generally curved forward in front of the column and sensitive, the other curved round near its base and functionless. To this section belongs C. macrocarpum, Rich., on which the genus was founded, and some nine or ten others, at least. The following species, and perhaps a few others, belong to the group :— C. arratum, Lindl. C. Gromus, Linden § Heichb. f. V See preceding list. C. MAOROCARPUM, Rich. SEXUAL FORMS OF CATASETUM. 223 C. MACROGLOSSUM, Reichb. f. C. MACULATUM, Kunth (C. integerrimum, Hook., C. Wailesii, Hook.). C. Naso, Lindl. C. GEnsTEDII, Reichd. f. C. PLANICEPS, Lindl. (C. recurvatum, Link, Klotzsch, & Otto). C. PURUM, Nees (C. semiapertum, Hook.). | See preceding list. C. vrRIDIFLAYUM, Hook. (C. serratum, Lindl.). In some of the species one of the antenne stands forward in front of the other, and the hinder one is usually, perhaps in- variably, functionless, as pointed out by Darwin in similar in- stances; in others both are in the same plane and equally sensitive. The same character appears in certain species of the succeeding section, so that the superior lip of the male flower is the essential distinguishing character of the vresent group. Section II. Myanruvs.—Diecious ; lip inferior in the male, generally much flatter than in Eucatasetum, superior in the female, and galeate ; rostellum prolonged below into a pair of slender antenne, which are either in different planes, and one of them functionless, as in Eucatasetum, or both stand forward in the same plane, over the lip, and are equally sensitive. It is a large group, comprising between thirty and forty described species. The position of the antennes appears to offer the best primary character for its subdivision, after which the shape of the lip serves to distinguish two or three fairly natural groups of species. Antenne in different planes, one only sensitive. C. PILEATUM, Reichb.f. See preceding list. C. Curistyanum, Reichb. fe C. Laminatum, Lindl. C. saccATUM, Lindl. C. TABULARE, Lindl. R2 224. MR. R. A. ROLFE ON THE Antenne in same plane, both sensitive. à . BARBATUM, Lindl. Q . CALLOSUM, Lindl. . CERNUUM, Reichb. f. Q | See preceding list. Q . CRISTATUM, Lindl. Q . DARWINIANUM, Rolfe. . DELTOIDEUM, Lindl. . GARNETTIANUM, Rolfe. . LURIDUM, Lindl. . SANGUINEUM, Lindl. See preceding list. . TRULIA, Lindl. CO OO OC Both these groups, but especially the latter, contain other species, but in most cases the material to hand is insufficient for satisfactory determination. Section III. EcrgRosz.—Diccious; lip inferior in the male, as in Myanthus, but the column without the antenne-like pro- longations of the rostellum ; female at present unknown. A remarkable little group of about eight described species, as follows :— C. CALCEOLATUM, Lem. C. GLAUCOGLOSSUM, Reichb. f. C. RosEUM, Reichb. f. C. RussELLIANUM, Hook. C. scu ERA, Reichb. f. . C. THYLACIOCHILUM, Lem. C. vIOLAScENS, Reichb. f. et Warscew. C. Warscewiczu, Lindl. Section IV. PskvDo-CarasETUM.— Hermaphrodite ; lip inferior, saccate ; column short and stout, without cirrhi. This group is remarkable on account of its hermaphrodite flowers, which, however, do not appear to differ in any other essential point of structure; while the habit. of the species 18 Role. Linn. Soc. Journ. Bor. Vor. XXVILP].B. M.Smith del. cm pie ss | RMorgan lith. ; s West, Newman imp. SEXUAL FORMS OF CATASETUM. 225 precisely identicalwith those of the preceding sections. Although I was able to examine living flowers, I could not satisfy myself whether any portion of the column was sensitive. I touched it in several places, somewhat roughly at last, but without producing the slightest effect. But when at length I brought the anther-case away by force the stipes of the pollinia was seen to be curved round a protuberance on the face of the column, and being liber- ated it immediately straightened itself; and I suspect that when the flowers are in the proper state of maturity the function of this elasticity is to help in ejecting the pollinia. Here is a case for future observation ; but I have noticed in one or two instances in the preceding sections that the flowers are not equally sensitive at alltimes. Whether this depends upon the proper stage of maturity being reached, or upon the temperature, or whether it is that some species are more sensitive than others, I do not at present feel satisfied about. Further observations are necessary to decide the point. The following species belong to this section :— C. cassrpEUM, Linden & Reichb. f. C. piscotor, Lindl. (Monachanthus discolor, Lindl., M. fim- briatus, Gardn., M. Bushnani, Hook., C. roseo-album, Hook.). C. LONGIFOLIUM, Lindl. DESCRIPTION OF PLATE VIII. Fig.1. Plant, showing the male and female inflorescences on either side of the same pseudo-bulb ; 3 nat. size. - À portion of male inflorescence; nat. size. - Lip of male; x 2 diam. : - Column of ditto, showing the antenn ` X 2 diam. . Anther-case, front view; X 3 diam. Pollinia, with stipes and gland, back view; x 2 diam. - The same, front view ; x 2 diam. - À portion of female inflorescence ; nat. size. TU o4 CO Lë Snows 226 DR. M. T. MASTERS ON THE MORPHOLOGY, Review of some Points in the Comparative Morphology, Anatémy, and Life-History of the Conifere. By MaxwELL T. MASTERS, M.D., F.R.S., V.P.L.S., Corr. Memb. Instit. France. [Read 18th April, 1889.] INTRODUCTION. Tue following review is the outcome of personal observation and research carried on at intervals during several years. The observations have, for the most part, been made upon living plants and fresh specimens. The living examples have been studied in the Royal Gardens, Kew, and in numerous pineta, gardens, nurseries, and plantations throughout the country. Moreover, I have myself had under cultivation from the seedling to the adolescent stage a large number of species and varieties, which have thus been under more or less continuous observation for some years. Numerous specimens have been obligingly sent to me from various botanic gardens, such as Edinburgh, Glasnevin, Cambridge, and from the rich experimental garden at Antibes presided over by M. Naudin. Native specimens have also been procured from Australia through the kindness of Baron Sir Ferdinand von Mueller, from New Zealand by the courtesy of Dr. T. Kirk and Mr. Hamilton, from America, from Spain, Portugal, Italy, and France, and some have been collected by myself in Switzerland. I am prevented by considerations of space alone from acknow- ledging separately the assistance I have received from proprietors, gardeners, and nurserymen. For much assistance in the matter of microscopical illustrations I am indebted to Mr. Nicholson and to Mr. Reed, and I have also availed myself of the excellent series of sections made to illustrate American Conifers by Dr. King. In addition to the fresh specimens I have made use of the materials in the London herbaria, and have controlled and ex- tended the observations made upon the living or cultivated plants by the comparison of the wild ones as represented in the herbaria. The illustrations are mostly from original sketches, and several are taken from the * Gardeners' Chronicle, to the proprietors of which Journal I am indebted for their use. ANATOMY, AND LIFE-HISTORY OF THE CONIFER. 227 The extensive literature,and especially the pictorial illustrations, of the order have also been systematically referred to. The main object of the investigation has been to gain a general and comparative view of the external morphology of the whole order, to ascertain, so far as possible, the causes and conditions inducing the development of particular forms or modes of growth, and to enquire into the purposes served by the numerous variations and presumed adaptations. The comparative histology of root, stem, and leaf is only in- cidentally alluded to, as these subjects are treated of in the works of Bertrand, De Bary, and Van Tieghem, which are fre- quently cited. Classification, or the study of the origin, lineage, and relation- ships of species and other groups, has also not been a primary object of investigation, although it has been constantly kept in view and the importance of certain characters as a means of dis- criminating between one species and another is duly noted. It may here be stated that I have followed, for the most part, the arrangement and limitations proposed by Bentham and Hooker and by Parlatore ; but, not being concerned in the present paper with the accurate nomenclature and synonymy of species, I have taken the names as 1 found them in gardens or in books, only correcting obvious errors, but without in all cases verifying (or having the means to verify) the assigned names. Various contested points in the morphology of the order, such as the nature of the needles of Sciadopitys and of Pinus, and the constitution of the flowers, male and female, have been dealt with in more or legs detail, but no attempt has been made to enter upon the question of Gymnospermy, the correctness of which is throughout assumed, nor to investigate the minute structure of the nucellus and its developments in the shape of archegonia and their relationship to similar produetions among the higher Cryp- togams. These matters are only incidentally alluded to, not, indeed, from any want of appreciation of their extreme i mportance in illustration of the lineage and descent of the plants in question, but from a personal lack of the requisite conditions for their efficient study. The “ characters " presented by Conifers, as by other plants, are partly the outcome of hereditary transmission and of survival in a comparatively unchanged condition. Partly also they are acquired or assumed in response to some influence from whose 228 DR. M. T. MASTERS ON THE MORPHOLOGY, power there is no escape short of gradual extinction. These latter are adaptive and serve, or have served, some useful pur- pose. Thus the nature of the root depends upon the conditions of the medium in which it grows; the direction of the branches, and the arrangement, structure, and movements of the leaves are clearly in relation with the varying intensity and changing direc- tion of the light. Other modifications have reference either to the protection or to the dispersal of the pollen or to similar requirements in the case of the embryo and seed. These and like matters are kept in view throughout the following pages. The observed changes are, as universally the case, brought about by variations in the intensity of growth and development, as by an arrest, excess, or perversion of growth and by similar conditions of development. Growth is here understood to be mere increase in bulk without attendant differentiation, while develop- ment implies progressive or regressive change—greater or less differentiation. Changes in the one process are often, but not necessarily always, associated with corresponding diversity in the other. Any particular change in one direction is pretty sure to be accompanied by co-relative changes in others. The changes in question may be continuous and uniform or they may be interrupted fora time to be afterwards resumed. These several mutations and rhythmical alternations bring about manifold or pleiomorphie conditions, which in the case of the Conifere are very remarkable. One such illustration may here be given in the case of the shoots of Pinus from the bud-stage to the fully developed state. In the unopened bud we have an illustration of arrested growth and development in the case of the bud-scales ; the central axis is also for the time arrested in its growth, the arrest being followed, in response to the stimulus of augmented heat, by rapid elongation, and by the corresponding development of the secondary leaves. In the male inflorescence there is also arrested development manifested in the enveloping bud-scales, followed in due course by the elongation of the axis and the differentiation of the tissues of the leaves into sporophylls or anthers. In the female flower there is similar arrested deve- lopment in the environing scales, and generally speaking 4 relatively low degree of differentiation in the bracts. These are serially continuous with, and obviously homologous with the leaves. The fruit-scales within the bracts present some remarkable illustrations both of arrest and, so to speak, . of ANATOMY, AND LIFE-IISTORY OF THE CONIFERS. 229 perverted differentiation and which are alluded to under their respective headings. © To show the diversity in appearance which may be met with in one member—the leaf—Engelmann points out that in Pinus seven different forms of leaves occur, as hereafter mentioned in detail. Certain of these forms of leaf occur on the same plant at dif- ferent periods, being characteristic either of youth or of tem- porary and intermittent accessions or relaxations of growth. In other cases a certain degree of permanence of a particular form of branch or leaf is observable. Mention is made of these cases under the head of ramification and foliage, and especially with reference to the Serpent-firs (Abies excelsa, vars. monstrosa and viminalis) and to the Retinosporas of gardeus. The Retinosporas have been considered to form a distinct genus, and are so described in various works of authority. They repre- sent, however, only various “ stadia” in the life-history of certain species of Thuya, of Chamecyparis, or of Juniperus. These plants, under cultivation, continue year after year to produce one de- scription of foliage only, and so may very readily be mistaken for separate species. Thuya pisifera thus exists under three or even four distinct conditions ; and if the clue were not afforded by some “ happy accident," such as the occasional presence of two or more forms on the same bush, all four forms might be considered as so many species, a fact not to be lost sight of in the identification of fossil plants. The forms in question are, as has been said, more or less persistent. Why, then, under favourable conditions should they not remain completely so? May we not, in fact, have, as Caspary suggested in the ease of the Snake-fir, the beginning of a new species? And may we not, especially in such instances as the Retinosporas, be lookers-on at the origin of forms which will gradually become stable enough to eonstitute species? The supposition is rendered the more plausible in that the variations in question are in some cases perpetuated by seed. The larger proportion of seedlings from such plants doubtless revert to the typieal stock ; but a considerable proportion per- petuate the variety. The cultivator, by the constant elimination of some forms, and by the continuous protection of others, suc- ceeds at length in “fixing " certain desired varieties. Doubtless a similar process occurs in Nature; and in watching the morphological changes that occur in plants, we are certainly Witnesses, on the one hand, of the permanence characteristic 230 DR. M. T. MASTERS ON THE MORPHOLOGY, of hereditary endowment or of reversion to an ancestral state, and, on the other, of the changes resulting in adaptation to vary- ing circumstances ; and, lastly, we may be present as on-lookers, and even as gardeners are sometimes privileged to be actual as- sistants-at the birth and development of new species. Whether, however, the phenomena witnessed in the so-called Retinosporas are survivals and vestiges of a former condition, or whether they are the results of attempts to conform to new conditions, which will ultimately result in “ new species,” is a matter that cannot be stated with any certainty. Those appearances in the adult plant which reproduce the characteristics of the seedling may be due to reversion, but others may be instances of progressive adaptation. Tur SEEDLING PLANT. The characteristics presented in the process of germination and by the seedling plant during its growth do not differ mate- rially from those of flowering plants generally. Nevertheless, there is a good deal of variation in minor points in different genera. Thus it is, in general, easy to distinguish a seedling Abies from a seedling Picea, and both from a Pinus; hence the desirability, for purposes of discrimination, of taking note of the differences which the seedling plants present. The investigation of the characteristics of seedlings derives further interest from the relation they bear to the supposed ancestral condition of the plants, from their bearing on the varying physiological requirements of the young plant, and on their power of adaptation to particular external circumstances. It is, however, requisite to exercise much cauiion in these matters, as it occasionally happens that in different species of the same genus different modes of germination occur, as, for instance, iu the genus Araucaria, where, in some species, the cotyledons are hypogeal and with a fleshy tigellum, epigeal in others. The lobing of the cotyledons in Pinus or the variable number of the seed-leaves may be indications of varying degrees of energy of growth, rather than of direct hereditary transmission. Grrmination.—The germination of the seed presents no very special features ; the husk of the seed splits into two pieces or irregularly, the radicle protrudes, descends into the soil, while the tigellum or caulicle lengthens often to a considerable extent, bearing at its summit the cotyledons still enclosed within the husk till at length, by the arching of the upper end of the ANATOMY, AND LIFE-HISTORY OF THE CONIFERS. 231 tigellum, and by the tension exercised by the growing cotyledons, the husk is thrown off. The mode of germination is still unknown in many species. I have had the opportunity, thanks to the assistance of the Di- rector and the Curator of the Royal Gardens, Kew, of examining a large number of species; and for a general statement of the observed phenomena reference may be made to the works of L. C. Richard *, Geeppert t, and to the résumé to be found in Sachs's Text-book t, or Goebel $. An investigation of the characters of the seedling plant is im- portant not only as occasionally furnishing a means of discrimi- nation, but also as showing vestiges of what has been, and as affording forecasts of what will be. The Radicle.—The monopodial radicle is usually slender and tap-shaped, its branches being arranged in two ranks, and given off generally nearly at right angles to the main axis, even in cases where they subsequently descend. The fibrils are often contorted, even when the soil they penetrate is light and open. The Root.—There are considerable differences in the depth to which the tap-roots penetrate, and in the extent and direction of their ramifications, dependent mainly on the character of the soil and the other conditions under which the plants are growing ; though in other cases this direct relationship is not so obvious. Thus, in Pinus silvestris, austriaca, rigida, insignis, in Picea ex- celsa, Sequoia gigantea, Cedrus Deodara, and C. atlantica the main root descends more or less vertically, while the branches spread nearly horizontally. In Pinus Laricio, ponderosa, Abies Pinsapo, Pseudotsuga Douglasii, Thuya gigantea, and Cupressus Lawsoniana the secondary roots have a downward direction. In Tsuga Merten- siana there is less difference in size between the primary and the secondary root-branches than in the Pines; the root, in fact, is more bushy, and the secondary branches go obliquely downwards, In Taxodium distichum the primary root is relatively extremely long, bright red, with a very few branches which are directed obliquely downwards. As compared with the Deodar (Cedrus Deodara), Cedrus atlantica has a more compact, less-spreading root-system with the secondary branches shorter. The cortex of * L. C. Richard, * Comm, Bot. de Conif.' (1826). t Goppert, ‘De Conif. Structura Anatomica ' (1841). 1 Sachs’s * Text-book,’ ed. Vines (1882), p. 508. $ Goebel, * Outlines of Classification, &c.’ (1887), p. 220. 232 DR. M. T. MASTERS ON THE MORPHOLOGY, the radicle of the Deodar is of a grey colour, that of C. atlantica red. Whether these appearances, observed in a small rumber -of cases only, are invariable, may well be doubted. M. Van Tieghem draws attention to the means whereby the radicle of some Conifers (excluding-the Abietines) is strength- ened and enabled to resist the strains to which it is subjected *. This is effected by the formation of thickenings in the constituent - cells of one of the cortical layers next to the endoderm. These thickened cells are peculiar to the root and are not seen in the caulicle f. The Caulicle:—The cauliclet is generally erect, slender, of con- siderable length to raise the cotyledons above the surrounding herbage, and of course destitute of branches or rootlets. The colour of the bark of the adult tree is often foreshadowed in the colour of the rind of the caulicle. In Cephalotazus and in some species of Araucaria, as in A. Bidwillii and A. imbricata, the caulicle is carrot-shaped, very thick and fleshy, thus serving as a reservoir of nutriment for the young plant. The Cotyledons.—The most noteworthy points with reference to these organs are the hypo- and epigeal conditions respectively. Where the cotyledons do not appear above ground, or are not much raised above its surface, they are generally thick and fleshy in texture, and do not readily disengage themselves from the seed-coat (see figs. 1,2, 3). They contain much nutritive matter available for the growing plant. Cotyledons of this character are met with inthe tribe Taxe: in Cephalotaxus and Ginkgo, and in the section Columbea of the genus Araucaria. The cotyledons of Ginkgo are thick, fleshy, oblong, constricted at the base into short stalks which in germination protrude in an arching direction, leaving the body of the seed enclosed within the shell. The plumule * P. Van Tieghem in Ann. Sc. Nat. sér. 7, tom. vii. p. 374 (1888), ** Sur le ré- seau de soutien de l'écorce de la racine 77 and Ann. Se. Nat. (1871), t. xiii. p. 187. + P. A. Dangeard, in a paper published since this communication was-read, and entitled * La mode d'union de la tige, et de la racine chez les Gymnospermes," Comptes Rendus, Feb. 1890, asserts that if there are two cotyledons there are two woody bundles in the root alternating with two liber-bundles; if three eotyledons, the number of bundles corresponds; but if there isa greater number of eotyledons, the number of bundles in the root does not correspond. f The word “hypocotyl” has of late years been introduced to denote what was formerly called the caulicle or the tigellum ; but as there appears to be no special advantage in departing from the rule of priority, and in adding a synonym to an overburdened terminology, I adhere to the old usage. ANATOMY, AND LIFE-HISTORY OF THE CONIFERS. 233 above the cotyledons is 3-sided, bearing tristichous leaves of an oblong form, gradually passing into the wedge-shaped form which characterizes the adult plant. Fig. 1.—Seedling plant of Ginkgo, drawn by Mr. Weathers. The mode of germination in different species of Araucaria Varies so much that authors have founded subdivisions of the genus upon these differences; thus in the American species belonging to the section Columbea there are but two cotyledons, and these hypogeal, ‘as in A. Bidwillii and A. imbricata. In 234 DR. M. T. MASTERS ON THE MORPHOLOGY, these species the caulicle is thick and fusiform. In the section Eutassa the cotyledons are four in number, leafy and epigeal, and ——— = — i za Fig. 2.—Germination of Araucaria imbricata; early stage, cotyledons enclosed within the seed, plumule commencing to protrude. Fig. 3.—Germinating plant of Araucaria imbricata, in a further stage of deve- lopment, showing hypogeal cotyledons, erect plumule, and primordial leaves similar in form and arrangement to the adult ones. are raised upon a well-marked caulicle, as in A. excelsa, A. Cookii, A, Cunning hami and A. Cuninghami glauca. Bentham, however, remarks that the number of cotyledons varies in the same species M We have here, then, illustrations of the fact that very conspicuous characters, and such, as in the case of the hypo- and epigeal cotyledons, are of great physiological significance, may be of rela- tively little value for classificatory purposes, inasmuch as species of the same genus may present variations in these respects. In those cases where the cotyledons are epigeal and leafy in ‘character, the chief points of distinction lie in their number, which may be two or more. Species with two cotyledons are usually very constant in this respect, it being very rare in such plants to * Bentham in Benth. et Hook. f. Gen. Plant. iii. p. 437. z E A - ANATOMY, AND LIFE-HISTORY OF THE CONIFER X. 235 find seedlings with more than two; but when the number ip increased, much irregularity as to number prevails. Duchartre’ , adopting the opinion of Adanson t and of A. L. de Jussieu 1, in opposition to the views of Gaertner $, Salisbury |, and L. C. Richard J, considered that the multiplicity of cotyledons was more apparent than real, and might be traced to the subdivision of two. If, however, the vascular bundles be traced from the caulicle, it will be seen that the vascular cylinder breaks up, not first into two divisions which subsequently branch, but into a variable number, not always in direct relation to the number of the cotyledons. The following genera and subgenera are known to have two M more cotyledons. Those genera to which the sign ! is appende have been examined by myself. Embryo dicotyledonous. Tribe Cupressinee. l Callitris !, Frenela !, Actinostrobus, Fitzroya, Libocedrus, Thuya L ~ Cupressus ! “(including C. Lawsoniana IA Juniperus !, Retino- spora 1 T Tribe Taxodiee. l Cryptomeria !, Taxodium ?, Athrotaxis !, Sciadopitys !. Tribe Tavee. Taxus !, Torreya!, Ginkgo !, Cephalotaxus!, Phyllocladus!, Dacr - dium !, Pherospheera. Tribe Podocarpee. Microcachrys, Saxe-Gothiea, Podocarpus ! Tribe Araucariee, ` Cunninghamia, Agathis !, Araucaria ! \ Embryo polycotyledonous. Tribe Cupressinee. ` Callitris spp. !, Cupressus spp. !, Juniperus spp. ` Tribe Taxodiee. Sequoia, Cryptomeria (subinde) ! Tribe Araucariee. Araucaria spp. ! * Duchartre in Aun. Ge, Nat. sér. 3 (1848), x. p. 207. Tédon: + Adanson, ‘ Familles, p. 805, “ le pin n’a réellement que deux cotylédons. * A. L. de Jussieu, * Gen. Plant.’ ed. 2, Pref. p. xviii, & p. 415. $ Gaertner, ‘De Fruct. et Sem.’ p. 157 (1788). || Salisbury in Trans. Linn. Soc. viii. (1807), p. 311. T Richard, Comm. Bot. de Conif. (1826). Hi 236 DR. M. T. MASTERS ON THE MORPHOLOGY, Tribe Abietinec. l Pinus !, Cedrus!, Picea (Link) |, Tsuga !, Pseudotsuga ', Abies ~ (Link) !, Larix ! . The following list, in which are incorporated the records of Goeppert and of Engelmann, together with my own, will show how variable is the number of cotyledons when more than two, even in different individuals of the same species. Abies apollinis 7, balsamea 4-5, cephalonica 4-6, graudis 6, pectinata 3-7, nobilis 6-8, Pinsapo 6-7, sachalinensis 4-5, sibirica 3-7, Web- biana 4, Veitchii 4-5. Araucaria spp. 2-4. Callitris sp. 2-4. Cedrus l atlantica 9, Deodara 11, Libani 6-9-11. Cryptomeria japonica 2-3. Cunninghamia sinensis 3. Cupressus glauca 3-4-6, globularis (Hort. Palermo) 2, pendula 3, pyramidalis 3, retrofracta (?) 2. Keteleeria Davidiana 2. Larix europæa 5-7, Griffithii 5-7, microcarpa 4-5, sibirica 4-5. Picea excelsa 3-8, Menziesii 5-6, orientalis 9-11. Pinus aristata 6-8, australis 7-10, austriaca 3-6-8, Ayacuite 12-15, Balfour- iana 5, Banksiana 4-5, Bungeana 11, calabrica 8, canariensis 6-7-8, Cembra 8-14, cembroides 9-12, cephalonica 5-8, cilicica 8, Coulteri 10-14, edulis 7-10, Elliottii 6-9, excelsa 8-12, flexilis 8-9, Gerar- diana 3-6-8, glabra 5-6, halepensis, 6, 7, 8, 9, inops 4-6, insignis 6-9, Jeffreyi 10, Lambertiana 12-15, Laricio 4, 6-8, maderensis (2) 10, maritima 6-8, Massoniana (?) 6, mitis 4-7, monophylla 7-10, monspeliensis 7-9, monticola 6-9, montana 3-6, palustris 9, parvi- flora 8-10, Parryana 8, Peuke 9-10, Pinaster 5-8, Pinea 8-14, Pithyusa 8, ponderosa 6-11, pungens 7-8, resinosa 6-7, rigida 5, Sabiniana 12-18, silvestris 3-6-8, Strobus 7-14, Teda 5-8, Tor- reyana 13-14, tuberculata 5-8. Pseudotsuga Douglasii 5-6-7. ANATOMY, AND LIFE-HISTORY OF THE CONIFERE. 2937 Sequoia gigantea 4-5. Tsuga canadensis 4. It is possible that an increased number of cotyledons might, under certain circumstances, be advantageous by securing a larger surface and a better chance in the competition with neighbouring herbage. The size of the cotyledons also varies greatly in different species. In Pinus_Pinea, for instance, they are 2 inches long and proportionately stout; while in P. canariensis they are equally long, but slender. The relative length of the cotyledons and of the primordial leaves which follow them may afford diag- nostic characters; thus, in some species of Abies, Pinus, Laris, Cedrus, and Athrotaxis the cotyledons are longer than the prim- ordial leaves, while in some species of Thuya, Cupressus, &c. the cotyledons are of about the same length as the succeeding leaves. In form the cotyledons are usually linear or linear-oblong, rarely broader as in Ketinospora (i. e. Thuya) obtusa, pisifera, Cryptomeria Lobbi, and Thuya gigantea. They are flat or rounded on both surfaces. In some cases the midrib is not pro- minent; in others, as in Abies Veitchii, sachalinensis, balsamea, cephalonica, and Apollinis, the midrib is prominent on the upper surface. In many of the species of Pinus the cotyledons are 3-sided, one side being anterior or inferior, the other two lateral às regards the axis. The margins are usually entire; but they may be hairy, and in Pinus Strobus they are distinctly toothed at the margins. Sometimes they are acute, or even mucronate at the apex, as in many species of Pinus; while in Abies they are Sometimes obtuse and even emarginate, as in Abies balsamea, sibirica, and Webbiana. In Picea excelsa they are thickened at the base, and thus afford indications of the peculiar “ pul- Vini" that are so characteristic of the adult branches in that genus. Anatomy.—A few incidental remarks may here be given, with a view to call attention to the importance of the subject in ita relation to physiology, and, as will be more fully exemplified in the case of the leaves, to the investigation of natural affinities, The fibro-vascular axis of the caulicle forms a cylinder between LINN. JOURN.—BOTANY, VOL. XXVII. 8 238 DR. M. T. MASTERS ON THE MORPHOLOGY, the cortical and the central cellular tissue, and is usually undivided for its whole length, breaking up only at the point of emergence of the cotyledons to send branches to those organs. In Arau- caria the vascular cylinder of the root is unbroken ; but in the fleshy tap-shaped caulicle it divides into four wedge-shaped masses with the xylem pointing towards the central pith. A section made at the point of emergence of the cotyledons shows four median bundles at right angles to the cotyledons. Above the cotyledons the bundles are increased to six disposed in a circle, while higher still they form a continuous cylinder. ‘The parenchyma of the ground-tissue ia the caulicle is crowded with starch-grains of different sizes; and betweeu the cortex aud the vascular ring it is traversed by a number of resin-canals disposed in a circle, or rather ellipse, around the vascular ring; another circle of resin-canals lies just beneath the cortex. In a species of Cupressus with rather thick cotyledons, the structure comprised one layer of brick-shaped epidermal cells, but no hypoderm. Beneath the epidermis on the upper surface only a single layer of palisade-cells was seen densely filled with chlorophyll. The central tissue consisted of loose parenchyma with spheroidal cells, for the most part destitute of chlorophyll. The central fibro-vascular bundle was, in section, transversely oblong, with a well-marked endoderm enclosing the xylem and the phloem, the latter directed, as in perfect leaves, towards the lower surface, the xylem in the opposite direction. In Pinus canariensis the three-sided cotyledons have a papular epithelium, no hypoderm, a mesophyll of spheroid or polygonal cells of uni- form character, but differing in size, the inner ones being the largest. There is a single fibro-vascular bundle surrounded by an imperfectly developed endoderm. The structure of the coty- ledon of P. Pinea is essentially the same. The peculiarities seem to reside in the papular epidermis (epithelium), the absence or imperfect development of hypoderm, and the absence of the sinuous cells which are so marked a feature of the mesophyll of the perfect leaves of the true Pines. Lestiboudois, in his * Phyllotaxie Anatomique,’ points out the frequent want of concordance between the number of the vascular bundles of the tigellum and the number of the coty- ledons. Although there are numerous variations, the normal position of the bundles is in the intervals between the cotyledons. ANATOMY, AND LIFE-HISTORY OF THE CONIFERZ. 239 In Pinus the general rule is to have twice as many cotyledons as there are vascular bundles in the caulicle, two cotyledons being placed in the intervals between the primary bundles: thus, sup- posing there are five bundles, as in Pinus Pinea, each bundle divides into two, each division extending into a cotyledon. P. canariensis, according to Lestiboudois, has four bundles and eight cotyledons, P. Strobus three bundles and nine cotyledons, P. palustris seven bundles and. nine cotyledons, P. maritima five bundles and six to nine cotyledons, P. Laricio four bundles and six cotyledons, or five bundles and six to eight cotyledons, P.`calabrica five bundles and eight cotyledons, P. monspeliensis four bundles and seven to nine cotyledons, P. excelsa five bundles and twelve cotyledons; Cedrus four bundles and eleven coty- ledon, the variations in number being due either to subdivision of the bundles, or to the opposite condition of inseparation *. In some cases where there are two cotyledons there are, ac- cording to Van Tieghem, two fibro-vascular bundles in the central cylinder of the root, each of which ramifies indefinitely in the median plane of the cotyledon, while in the polycotyledonous genera the number of fibro-vascular bundles varies even in dif- ferent individuals of the same species without, however, corre- sponding to the number of the cotyledons t. , The stomata on the cotyledons vary greatly in number; they are usually oval, with two-guard cells. As to their position, it 18 noteworthy that they often occur on the upper surface, or on thelateral faces where the cotyledons are three-sided, even in Cases where in the adult leaves they are placed on the lower Surface. In many species of Pinus and in Picea Menziesii, where the cotyledons are three-sided, the stomata are almost exclusively on the two lateral surfaces, the dorsal or anterior surface being nearly ; if not quite, destitute of them. The same relative posi- tion of the stomata occurs in the adult leaves of these plants in Some Junipers, Picea ajanensis, &c. * Lestiboudois in Ann. Sc. Nat. sér. 3 (1848), p. 23, t. x. t Van Tieghem, * Traité de Botanique'(1884), p. 1323. For fuller details consult De Bary, ‘Comparative Anatomy of the Vegetative Organs of Phanerogams and Ferns,’ ed. Bower and Scott (1884), pp. 245, 356, 386. Goppert, * De seminum germinatione in De Coniferarum Structura Anatomica’ (1841). Lately Messrs. Van Tieghem and Douliot have shown that in all Gym- nosperms, as in Angiosperms, the radicels come wholly from the pericycle, and are furnished with a root-cap. 82 240 DR. M. T. MASTERS ON THE MORPHOLOGY, This position of the stomates on the upper or lateral surfaces of the cotyledons has been specially observed by the writer in ‘Pseudotsuga Douglasii, Abies grandis, balsamea, cephalonica, Apollinis, sachalinensis, sibirica, Veitchii, Cedrus atlantica, Libani, Oryptomeria Lobbi, Cupressus Lawsoniana, sempervirens, funebris, Pinus Jeffreyi, Retinospora pisifera, and Thuya gigantea. Engel- mann * notes that in Sciadopitys the stomates are placed on the lower surface of the cotyledons exclusively. Plumule.—The term plumule is not very accurately defined. For our present purpose we may take it to be that portion of the axis with its attendant leaves which is formed in the embryo plant prior to germination above the cotyledons. Sometimes it remains in the same condition as when first formed ; but gene- rally it undergoes changes in size, &c., during and after germi- nation. The leaves which it produces are often different from those formed on the other parts of the stem. Growth in the plumule is for a time continuous, or at least relatively so; so that sometimes no definite limit can be observed between it and the succeeding portion of the axis. Atother times plumular growth is arrested by the formation of a terminal bud invested by perular scales, as I have observed in Pinus silvestris, Pseudotsuga Dou- glasii, Abies pectinata, sibirica, cephalonica, Picea orientalis, and P. Menziesii. How far this is a congenital character, or to what extent it is due to external conditions, is uncertain. In Gingko biloba there is a gradual transition between the perule and petiolar expansions to perfect leaves. Sometimes the axis above thecotyledons lengthens before it produces any leaves, and then, of course, the plumular or primordial leaves are removed from the cotyledons by internodes. In other cases, as in Cephalotarus Fortunei, Hetinospora pisifera, R. obtusa, Thuya gigantea, Cu- pressus sempervirens, C. Lawsoniana, C. funebris, and C. macro- carpa, the first two leaves are close to the cotyledons and decus- sate with them, while the subsequent pairs of leaves may be separated by internodes. In other cases, again, the primordial leaves are so crowded that they may be called tufted, as has been observed by the writer in Frenela, Lariz, various species of Pinus, Retinospora obtusa, Picea orientalis, Thuya gigantea, Pseudotsuga Douglasii, Abies cephalonica, grandis, sibirica, and Apollinis species in which, as before stated, a true scaly winter- * " Revision of the Genus Pinus,’ p. 3. ANATOMY, AND LIFE-HISTORY OF THE CONIFERS. 241 bud marks the termination of the first stage of growth. When the axis lengthens, the primordial leaves, instead of being tufted are either scattered as in Sequoia gigantea, where they are alter- nate, or arranged in decussate pairs, each pair being separated by an internode from that above and that below, as in Scia- dopitys, Taxus, Ginkgo, Cupressus, Callitris, Frenela, Araucaria spp., Libocedrus, Cedrus, Phyllocladus, Thuya, Picea Menziesit, and Cryptomeria. In some of these plants axillary buds may be seen in the axils of the primordial leaves, the first two leaves of these buds being placed at right angles to the primordial leaf. The upper part of the plumule in Pinus Torreyana and Retino- spora pisifera is conoid, compressed from front to back, and depressed at the apex owing to the projection on either side of a leaf-tubercle *, . In a specimen of Cephalotaxus Fortunei in the Kew Museum, above the cotyledons and in close proximity to them is a pair of Opposite leaves followed after an interval by a whorl of four linear leaves rather larger than the preceding, then after another inter- Space occur two leaves succeeded in a decussate manner by two others. In this plant the plumule emerges from the seed before the cotyledons are disengaged from the seed-coat. The primor- dial leaves are either free at the base, as in Retinospora obtusa, &c., or concrescent t with the stem and uplifted with it, as in Cupressus sempervirens, Abies Veitchii, and Oryptomeria japonica. l These observations relating to the seedling plant have necessarily been made on a limited number of examples; care must there- fore be exercised in drawing inferences from them, the more so 48 some of the peculiarities mentioned are manifestations of relative degrees of vigour dependent on external conditions $. . * Dingler, as cited in the * Botanical Gazette,’ May 1883, p. 229, asserta that in seedling Gymnosperms there is, as in Cryptogams, a single apical cell, instead of a group of cells at the apex of the growing-point as happens in Flowering Plants, but this needs confirmation. t The word “ conerescent " is adopted by M. Van Tieghem, and seems to be coming into general use to express the condition formerly, but. erroneously, ed adnation, and by myself * inseparation,” as denoting that kind of arrest ` a developmental change indicated by imperfect separation of two ports. The “rrest of development in these cases is associated with rapidity and vigour of growth d It may be convenient to append in this place references to various works in which the mode of germination of different species is figured :— balsamea, pectinata, Pinsapo, Duchartre in Aun. Sc. Nat. sér. 3, t. x. 242 DR. M. T. MASTERS ON THE MORPHOLOGY, ForraTrON. The leaves of Conifers assume varied appearances according to the age of the plant, their position and office. Thus, passing from the seedling state with its cotyledons already noticed, there may be primary leaves, secondary or true foliage-leaves, bud-scales or perulæ, bracts on the male and female inflorescences and anther- bearing leaves. These will be mentioned under their respective headings *. Comparatively rarely all three portions of a perfect foliage- leaf are present at the same time. The blade, or lamina, is the part most often wanting ; ; indeed it is questionable whether what is so called is not, in most cases, an expansion of the petiole, destitute of those modifications of venation and nerve-endings which characterize the lamina proper. In Ginkgo, for instance, the stalk expands into a leafy blade without any true midrib, the numerous veins being of nearly equal size, diverging from the top of the stalk-forking as in Ferns, but not connected one with another by side reticulations. The leaf-stalk, according to Fankhauser and Thomas, has the same anatomical structure as the secondary leaves of Pinus hereafter mentioned t. In the Yew (Tuzus), the Hemlock Spruces (Tsuga), and some others, the petiole is well marked and distinct from the lamina, tab. 9 (1848). Araucaria excelsa (Dombeya), Lambert, Pinus, ed. major, t. 39; Forbes, Pinetum Woburnense. Abies pectinata, Willkomm, Forstl. Flora, p. 113. Cedrus Libani, L. C. Richard, Comm. Bot. de Conif. t. 14 ; Duchartre, Le, Dacrydium, Nouv. Arch. Mus. iv. t. 3. Ginkgo, Le Maout and Decaisne, Gen. Syst. ed. Hooker, p. 747. Larix europea, Duchartre, l. c. ; L. C. Richard, l. c.t. 13. Picea excelsa (Link), Richard, l. c. t. 15; Willkomm, Forst]. Flora, p. 59; Fischer, Beitráge Monocot. und Polycot.; Sachs, Text-book, ed. Vines (1882), p. 508. Pinus Cembra, Fischer, Beiträge. P. Elliottii, Engelmann, Revision Pinus. P. Pinea, L. C. Richard, l. e. P. Salzmanni, Dunal in Mém. Acad. Se. Montpellier, ii. p. 15, t. 2. P. silvestris, Willkomm, Forstl. Flora, p.103. P. pumilio, Nees, Gen. Plant. Flor. Germ. P. Strobus, Henry in Act. Acad. Cæs. Leopold.-Carol. Nat. Cur. xix. p. 1 (1837), tab. xii. fig. 2. Pinus australis, excelsa, Laricio, monspeliensis, Pinea, Pinaster, Duchartre in Ann. Sc. Nat. loc. cit. Taxus, Nees, Gen. Plant. Flor. Germ. Various species of Phyllo- cladus and Podocarpus are figured in Kirk's * Forest Flora of New Zealand.’ * See Zuccarini on the Morphology of the Conifers, section vi. p. 32. English Edition, Ray Society, 1846. + Fankhauser, ** Die Entwickelung des Stengels und des Blattes von Ginkgo biloba,” Bot. Centralblatt, 1882, p. 229. Thomas in Pringsheim's e Jahrbuch,’ iv. p. 24. ANATOMY, AND LIFE-HISTORY OF THE CONIFER X. 243 but the vaginal portion of the stalk is absent unless the “ pulvinus ” be taken to represent it, a matter discussed under the heading of the minute anatomy : suffice it here to say that no such pulvinus occurs in these plants at the base of the perule, whose vaginal character will not be contested. In the Red-Wood (Sequoia sempervirens), in Cunninghamia, in Cephalota«us, Torreya, Podocarpus, and others, the perulæ are also broad-based, sheath-like, and have no pulvinus. In some of these plants, moreover! the base of the laminar or laminoid portion is not contracted into any petiole, but is concrescent with the stem (decurrent). In many Conifers the leaf, in the adult state, is represented by the petiolar portions only ; but in other cases, as for instance in the bracts of the cone of Pseudotsuga Douglasii, there are in- dications of vaginal and petiolar portions. Arrangement of Leaves—Homotaxy and Heterotaxy.—The phyllotaxis of the leaves of Conifers is treated of in various text-books, so that a few peculiarities are all that demand atten- tion in this place. Sometimes, as in Sequoia gigantea, Araucaria imbricata, Abies Pinsapo, the foliage is arranged in the same way throughout the entire tree (Homotazy). More usually the Arrangement varies in different parts of the tree. A change in the phyllotaxy is of course common in the case of the bud-scales, and also in the passage from the primordial to the adult foliage : thus in young plants of Thuya gigantea the leaves may be seen in two, three, or in four ranks. In other cases the change is more apparent than real, being brought about by concrescence in various degrees, the concrescence alternating more or less regularly with dialysis or freedom. Again, on the erect “ leader- shoots” of various species of Abies, Picea, Pseudotsuga, Tsuga, Taxus, Cephalotaxus, &c., the leaves spread on all sides, while on the lateral branches, which spread more or less horizontally, they are arranged nearly in one horizontal plane * In such. cases the phyllotaxis is not altered, but the leaves are twisted More or less at the base so as to make them apparently di- stichous, while in reality they are still polystichous. Occasionally Some of the side shoots will suddenly quit the horizontal direction and assume an erect attitude, as in Picea Menziesii and P. ajanensis. When this happens, the leaves spread on all - Abies PP. insapo forms a constant exception to this rule, the leaves being in- variably spreading in all directions. 244 DR. M. T. MASTERS ON THE MORPHOLOGY, sides as in the true leader-shoots. This change of direction is the original of some of the “fastigiate” varieties, as of the common Yew, where the branches, instead of spreading more or Fig. 4.— Cephalotaxus Fortunei, leader-shoot with scattered many-ranked leaves. less horizontally, assume an erect position, and the leaves are then given off on all sides, not being twisted at the base (see figs. 4, 5). À fastigiate branch may thus originate as a sport or bud- variation, as is seen in Tazus, Cephalotaxus, and other species. A remarkable correlation may frequently be observed in such cases between the size and the direction of the leaves. Thus in the variety brevifolia of Pseudotsuga Douglasii the stature of the plant is relatively dwarf, and the leaves not only spread on all sides but are much smaller than in the normal state, where they. spread in one horizontal plane only. In other cases the leaves ANATOMY, AND LIFE-HISTORY OF THE CONIFER E. 215 are longer than usual. M. André describes (in the ‘ Illustration Horticole’ for 1870, p. 106) a variety of the common Silver Fir, Abies pectinata, in which, in conjunction with a regularly pyra- midal outline of the whole tree, the leaves were not pseudo- distichous but spreading on all sides. Similar phenomena may N " B ZA NZ : EY BN Fig. 9.— Cephalotaaus pedunculata, var. fastigiata, showing the terminal shoot with scattered pluriseriate leaves, and a lateral shoot with pseudo-distichous leaves. be seen in the dwarf varieties of the common Spruce, Picea ex- celsa var. pygmea, &c. In Picea excelsa var. monstrosa and var. viminalis a very curious condition occurs in which the leaves are DR. M. T. MASTERS ON THE MORPHOLOGY, 246 re SS LS O Se À 4 X. FEA Ee N MASS Wf WE g varying dispositions of the leaves in consequence of alternating rapidity and slowness of growth. Fig. 6.— Abies Nordmanniana, showin ANATOMY, AND LIFE-HISTORY OF THE CONIFERÆ. 247 homotaxic, but much longer than ordinary, and owing to the rarity with which lateral buds are formed the main branches are elongated and unbranched, suggestive of a resemblance to snakes. It is interesting to compare these forms with the arrangement in young plants of Araucaria imbricata. These curious firs will be spoken of again under the head of “ ramification.” Changes in the appearance and disposition of the leaves also occur as a result of alternations of growth and of arrest of growth in various degrees. Thus, when growth first begins in spring, it proceeds rapidly, and then may come a period of arrest and con- tracted growth. Thus we may have, at different parts of the same branch, long leaves arranged in a pseudo-distichous manner, and others much shorter and spreading on all sides, as shown in the accompanying illustration from Abies Nordmanniana (fig. 6). The genus Podocarpus presents marked illustrations of hetero- morphy and of heterotaxy in the foliage. The leaves on some of the quickly growing shoots are distichous, arranged in one flat, horizontal plane, the individual leaves being broadly linear and somewhat faleate, while on the older or less quickly growing shoots the leaves are polystichous, short, subulate, and appressed to the stem. Another instance of variation in the arrangement of leaves is often seen in Abies Nordmanniana, A. Pichta, A. amabilis, as also in Tsuga canadensis, &c. The leaves on the lateral and more or less horizontally spreading branches, though polystichous, m reality arrange themselves in three rows, one on either side of the branch (in which case the leaves are nearly at a right angle to the branch), and one in the median plane of the upper Surface (in which case the leaves are appressed along the branch parallel to its main axis). The median leaves are usually smaller than the lateral ones. This arrangement may be compared to that in the species of Selaginella, or to that in Lycopodium spectabile, L. volubile, &c., where the lateral leaves are relatively large and decurrent, while the median leaves (dorsal as well as ventral in these instances) are relatively small, free at the base and caducous. In Abies grandis the uppermost row of leaves is Dot arranged parallel to the axis, as in the instances just referred to, but at right angles in the ordinary way, the obstruction caused by the overlapping leaves being minimized by the smaller size of the upper leaves, and by the power which the leaves have of *alsing or depressing themselves according to circumstances. 248 DR. M. T. MASTERS ON THE MORPHOLOGY, The leaves on the erect shoots being crowded and overlapping one another more completely than elsewhere, are less favourably situate as regards access of light than those on the side shoots, in which, owing to the arrangements before mentioned, the in- ternecine competition is less severe, hence one reason for the speedier fall of the leaves from the erect shoots and the more crowded parts of the tree, and the comparatively naked inter- spaces between the whorls of branches. The disposition of the leaves of these plants indeed seems to be arranged to secure an equitable share of light and as equal exposure as possible to the whole mass of foliage, thus reducing competition among individual leaves to the greatest possible ex- tent. In some cases, the arrangement is such as to secure the exposure of the stomatiferous surface to the light and heat, asin cases where the leaves are appressed against the stem in such a way as to leave the stomate-bearing surface free *. The movements in the leaves of many Conifers during the season of active vegetation particularly, and which take place under the influence of sunlight, have presumably a similar object. In Picea ajanensis the stomata are on the upper surface of the leaf, and consequently appressed to the lateral branches, while the green dorsal surface is exposed to the light. The uplifting of the leaves, however, permits the stomatic surface to be exposed upon occasion T. Mr. Moggridge alludes to this subject in a note on the movements of the leaves of Pinus halepensis (‘Journal of Botany, Feb. 1, 1867). The leaves of this tree in warm sunny weather are fully separated, but if the sky become overcast they close partially ; the sirocco produces a similar but more marked effect, but in rain the leaves collapse, giving the tree a most me- lancholy aspect. A similar change is observable in most Pines in summer and winter respectively, at which latter season they are nearly parallel, not divergent. Probably the state of things in some cotyledons, as before meutioned, and in the primordial * E. Mer, * De l'influence de l'ombre et de la lumière sur la structure, l’orien- tation et la végétation des aiguilles d’ Abies excelsa," in Bull. Soc. Bot. France, t. xxii. p. 199, t. xxiv. p. 109, t. xxvi, p. 15, t. xxvii. p. 23, and t. xxx. (1883) p. 40; and Comptes Rendus, April 16, 1883. t Chatin, “Sur les mouvements périodiques des feuilles d'Abies Nordman- niana,” in Bull. Soc. Bot. France, xxiii. p. 103 (1876). Masters, “On the Re- lation between Morphology and Physiology in the Leaves of certain Conifers," in Journ. Linn. Soc., Bot. xvii. (1879), 547, &c. Regent ANATOMY, AND LIFE-HISTORY OF THE CONIFERS. 249 leaves of Juniperus, Thuya, &c., in which the stomata are more abundant and more freely exposed than in ordinary foliage-leaves, may similarly be connected with the need for rapid evaporation through the stomata. In any case the arrangement, the torsion, and the movements of leaves seem in some cases, and at some periods, to promote the exposure of the assimilating or green side of the leaf, and in other cases (or in the same cases at different periods) to facilitate the exposure to light and heat of the stomate- bearing surface. In all probability these differences are directly connected with the assimilating process on the one hand, and with the respiratory and exhaling processes on the other. Anatomy of the Leaf:—The complete discussion of anatomical details lies without the scope of the present communication. Information concerning them may be obtained from the ordinary text-books and, in particular, from the memoirs cited in the bibliographical list appended to this communication. Certain peculiarities of structure have, however, been proposed as adjuncts to the discrimination and classification of the species, and on that account demand some notice here. The epidermis covering the adult leaves is represented in the cotyledons of many species by a soft papular layer apparently destitute of cuticle. In both cases the epiderm is perforated with stomata whose dis- tribution and number afford useful aids in classification, while their physiological significance is obviously a matter of importance. In most cases (in all, so far as my own examination enables me to say) the cotyledons have the stomata arranged on their upper- most or innermost surface (see p. 239). In the adult leaves the stomata are variously disposed on one or both surfaces, as is illus- trated by the following list taken principally from Bertrand * :— Stomata on lower surface chiefly :—Taxus, Torreya, Cephalo- taxus, Ginkgo, Saxe-Gothea, Podocarpus, Prumnopitys, Cunning- mia, Agathis, Cedrus Libani, C. Deodara, Picea sitchensis, jezo- ensis, Tsuga canadensis, Brunoniana, and Sieboldi, Pseudotsuga Douglasii, Abies firma, cephalonica, pectinata, Pindrow, cilicica, bifida, bracteata, Veitchii, sibirica, homolepis, &c., &c. Stomata on upper surface chiefly :—Juniperus, Athrotaxis Gunnianal, Libocedrus tetragona!, Cedrus atlantica, Piceaajanensis, Tsuga Hookeriana, Abies nobilis, Davidiana, grandis, Regine- Ra also Hildebrand,“ Bau d. Spaltéffaungen der Coniferen,” Bot. Zeit. 1860, pin. 250 DR. M. T. MASTERS ON THE MORPHOLOGY, Amalia, numidica, Pinsapo 8, Fraseri, balsamea, Larix Lyallii, americana, &c., Ze, Stomata on both surfaces :—Fitzroya, Cryptomeria, Taxodium, Sequoia, Athrotaxis cupressoides, Dacrydium, some species of Podocarpus. Araucaria Cunninghami (four bands), excelsa (four bands of 5 rows each), Balanse (4 bands of 8 rows), Cookii (2 bands of 50 rows), Rulei (3 bands), montana (3 bands), Muelleri (3 bands), brasiliensis (65 rows on each side), imbricata (70 rows on each side), Bidwillii (90 rows on each side)” (Bertrand). Picea nigra, alba, Khutrow, polita. "Stomata on two surfaces, but absent on the lower surface:— Frenela, Widdingtonia, Libocedrus, Thuyopsis, Cupressus, Juniperus, Athrotaxis laxifolia and selaginoides, Pinus Strobus, excelsa, &c., &c. In some cases the stomata are disposed in longitudinal bands, as in species of Pinus, Abies, &c., &c., or they may be irregularly dispersed, or they may occur in patches in certain privileged localities, as in Juniperus, where they occur in a triangular patch on the upper surface at the base of the spreading leaves, e. g. in Juniperus drupacea. . The position of the stomata in Conifers is very generally indi- cated by the existence ofa glaucous bloom *, but thisis not always confined to the immediate vicinity of the stomata, but may, as in Picea pungens var. glauca and others, be distributed over the whole surface of the leaf. The Hypoderm.—Within or beneath the epiderm is usually to :x50- Fig. 7.— Pinus patula. Plan of transverse section of leaf, showing epiderm, two layers of hypoderm, three resin-canals in the centre of the leaf, endoderm, and pericycle. be found a layer, or sometimes more, of long, thick-walled strengthening-cells, either forming an unbroken sheet or perfo- ` * Francis Darwin, “On the Relation between the Bloom on Leaves and the Distribution of Stomata," Journ. Linn. Soc., Bot. vol. xxii. (1886) p. 99. me ANATOMY, AND LIFE-HISTORY OF THE CONIFERE. 251 rated by the stomata, thickened by additional layers in certain spots, as at the angles of the leaves, over the midrib, round the resin-canals, and sometimes broken up into thick detached islets, à remarkable instance of which is afforded in Pinus Coulteri, the leaves of which have triangular wedge-shaped masses of hypoderm projecting into the mesophyll. Stahl says that in Abies pectinata the hypoderm-cells are most abundant on the side of the leaf ex- posed to the sun, least so on the shady side. Messrs. Coulter and Rose point out the presence in certain of the North-American species of Pinus of a layer of relatively thin- walled cells between the epidermis and the strengthening-cells proper. This layer is indicated by them in the following species: — P. monticola, albicaulis, flexilis, reflexa, Strobus, Ayacuite, resinosa, contorta, muricata, Engelmanni, Coulteri, ponderosa, tuberculata, arizonica, and Montezume. The Mesophyll.—Within the hypoderm, and forming the chief substance of the leaf, is the mesophyll, generally wholly cellular, the constituent cells containing chlorophyll being uniform or nearly so, as in Picea, while in others there is a distinction into palisade-cells, on one side (Abies), or on both surfaces (Araucaria imbricata), bui in any case most conspicuously on the side most exposed to the light, and branching cells with lacunæ between them (Abies). In some species the cells appear to radiate from the central bundle (Picea, spp.). In Pinus they are remarkable for the involutions of their walls, causing flat projections into the interior of their cavities. In other cases a series of colourless cells spread laterally from the central bundle in the midst of the ehlorophy ll-containing cells of the mesophyll and constituting the transfusion-tissue. This may be seen in Keteleeria Fortunei!, Pseudo-larix Kempferi!, some species of Podocarpus, e.g. P. chilensis! Very large, branching, thick-walled cells like those of Welwitschia are found in Araucaria imbricata and other species of the order, The Resin-canals.—Traversing the mesophyll are the resin- canals, which demand notice from the constancy of their arrange- ment *. Their presence is, to some extent, variable; but when they do occur it is usually in definite numbers and in certain * In the stem they occur in the middle of the cortex, generally as a single Tow, but.in some species of Pinus where the cortex is very thick there are several rows. They also occur in some cases in the secondary wood, whilst in Ginkgo they are to be found in the pith. 252 DR. M. T. MASTERS ON THE MORPHOLOGY, well-defined situations. Thus in all the species of Tsuga there is one central resin-canal immediately beneath the central vascular bundle, and a similar arrangement prevails in Torreya, Cephalo- taxus, Podocarpus, Saxe-Gothea, Sciadopitys, Fitzroya, Cryptomeria, Sequoia, Taxodium, Juniperus, &c. In the species of Larix there are two canals, one at each angle of the leaf beneath the epiderm. A very common arrangement is that in which the canals are placed within the epiderm on the lower surface on either side of the midrib, or there may be several canals arranged all round the leaf at intervals beneath the epi- dermis, as, for instance, in Pinus Strobus, monophylla, Cembra, and in some species of Araucaria and Altingia. In other instances, as in Pinus Pinaster, mitis, inops, Teda, Lambertiana, Araucaria imbricata, Abies Veitchii, &c., &c., the resin-canals are in the centre of the mesophyll, while in Pinus pseudostrobus Ze, they are almost or quite in contact with the endoderm. In three-sided leaves of some species of Pinus, if there be but one canal, it is placed under the epidermis in the centre of one side, if three then there is one in each of the three angles of the leaf. Engelmann accordingly spoke of the ducts as peripheral, or subepidermal, parenchymatous, and internal, and grouped the species of Pinus in large measure according to the position of the ducts. The arrangement of the canals is sometimes different in the primordial and in the secondary leaves respectively, and in Pinus Lambertiana the resin-canals are sometimes subepidermal, some- times parenchymatous. The canals are sometimes separated from the mesophyll by a well-marked ring of thick-walled cells, often continuous with the hypodermal cells, but occurring sometimes around those ducts which are placed in the centre of the mesophyll ; thus the resin- canals of Pinus are :— 1. Entirely surrounded by “ strengthening-cells," e. g. Pinus silvestris, Laricio, Pinea, serotina, nigricans, Peuke, deflexa, pon- derosa. 2. Encompassed by thick-walled cells with thin-walled cells intermingled, e. g. Pinus densiflora, Coulteri, Torreyana. 3. Begirt by thin-walled cells only, e. g. Pinus maritima, hud- sonica, Strobus, pyrenaica, insignis, Jeffreyi, longifolia, canariensis, eacelsa, argentea, monticola * * See Möbius, in Jahrb. f. wissenech, Botanik, t. xvii, p. 263 (1885). ANATOMY, AND LIFE-HISTORY OF THE CONIFERS. 253 The Endoderm.—The mesophyll is limited on the inner side by the endoderm or bundle-sheath, consisting of a single row of oval cells, sometimes more or less thickened by collenchymatous deposit at the points of contact. The endoderm is usually very well marked in species of Pinus, Abies, and Picea, while in Phyl- locladus, Podocarpus, Dammara, Fitzroya, Juniperus, and Sciado- pitys it is generally ill-defined. The Pericycle.—The endoderm immediately encircles the peri- cycle, which is composed of parenchymatous tissue of varying amount, destitute of chlorophyll, but often containing starch [probably always at some stage of their existence]. These cells are usually thin-walled and small, but are often intermixed with a small number of relatively very large, thick-walled prosenchy- matous cells. In those cases where the fibro-vascular bundle remains unbranched the pericycle is cylindrical, circular in out- line as seen in section, e. g. Pinus Lambertiana, Pseudolarix Kempferi, Picea spp., &c., &c.; but if the bundle is double, 7. e. divided, then the pericycle is transversely oval or reniform in section. The Fibro-vascular Bundle.—The centre of the pericycle is occupied by the fibro-vascular bundle. In most cases there is but one such bundle (constituting the midrib), but in other cases, as in Araucaria, there are several such bundles traversing the leaf, each with its own pericycle and forming several ribs. The central bundle is usually undivided throughout its course, but sometimes it bifurcates, and then on a transverse section an Appearance is presented as of two bundles, often separated at the dorsal side by a wedge-shaped mass of thick-walled cells; so that the two divisions of the bundle, instead of being in one transverse plane ( ), are now inclined (“ N) one to the other. In Pinus rigida and some others the masses of xylem, instead of pointing upwards and inwards, point outwards (^^ ^ ), diverging from the base. The central bundle consists of phloem elements on the outer or dorsal side and xylem on the inner side. In Sciadopitysthe relative position of these elements is reversed. The phloem-cells are more thin-walled than the xylem-cells, the mass of which latter presents in section a wedge-shaped out- line, the base towards the phloem, the truncated apex towards the centre and upper side of the leaf. The xylem-cells are usually packed in regular series, so that on section the rows seem to LINN, JOURN.—BOTANY, VOL. XXVII. T 254 DR. M. T. MASTERS ON THE MORPHOLOGY, radiate towards the centre, and sometimes the rows are separated by medullary rays. The phloem-cells, on the other hand, are much less regularly disposed. The total quantity of xylem and phloem in relation to the pericycle and the relative amount of xylem and phloem in the same bundle vary very considerably in different species. A single unbranched bundle within one pericycle is met with in Fitzroya, Juniperus, Cryptomeria, Sequoia, Athrotaxis, Cepha- lotaxus, Taxus, Torreya, Ginkgo, Phyllocladus, Saxe-Gothea, Po- docarpus, Cunninghamia, Araucaria, Sciadopitys, some species of Pinus, as in P. albicaulis, flexilis, reflexa, Strobus, Ayacuite, monticola, Lambertiana, monophylla, edulis, cembroides, latisquama, Parryana, Balfouriana, aristata, Cembra, in Abies nobilis, magni- Zog, Keteleeria Fortunei, Pseudotsuga Douglasii, in species of Cedrus, Picea, Larix, Tsuga, &c. A branched bundle occurs in the cladodes of Sciadopitys, 1n Abies amabilis, grandis, Lowiana, concolor, lasiocarpa, Frasert, balsamea, sibirica, Veitchii, firma, including bifida, homolepis, Pindrow, Webbiana, pectinata, Nordmanniana, cilicica, cephalonica, Pinsapo, numidica &c., and in the following species of Pinus :— bracteata, religiosa, contorta, muricata, Engelmanni, Coultert, ponderosa, Montezuma, Torreyana, Jeffreyi, Pinaster, Teda, Sa- biniana, serotina, rigida, insignis, pungens, tuberculata, inops, clausa, mitis, glabra, Banksiana, palustris, cubensis, Laricio, and silvestris. In considering these characters as available for systematic purposes it may here be remarked that the arrangement of the stomata usually affords a good (that is a relatively little variable) character, that the presence and thickness of the hypoderm are less reliable, being dependent on climatic influences and exposure, that the characters afforded by the mesophyll are of great import- ance, though not worked out in all the species, that the position of the resin-canals is of more consequence than their number, while the undivided or branched condition of the vascular bundle, as pointed out by Messrs. Coulter and Rose, is a character of high value. Minute structural characters are, however, no more absolute than other diagnostic points, although serviceable “ for use in the absence of other characters" *. * In addition to the authors incidentally mentioned and to the general treatises of Strasburger, Sachs, De Bary, Van Tieghem, &c., reference may here ANATOMY, AND LIFE-HISTORY OF THE CONIFER. 255 Homomorphy and Heteromorphy. The adult leaves, or those produced on the vegetative system of the plant, in succession to the cotyledons, may be of the same pattern or they may vary in different stages of the plant’s growth, or on different parts of its branches. In that form of Thuya known in gardens as T. Bodmeri, and in that known as Retino- suitably be made to others who have studied the anatomy of the leaves from various points of view :— Thomas in Pringsheim's Jahrb. iv. p. 43 (1865). Frank, “ Ueber den Einfluss des Lichtes," in Pringsheim's Jahrbuch, ix. p.147. Bertrand, Anat. Comp. in Ann. Se. Nat. sér. 5, vol. xx. p. 5, c. tabb. 12 (1874). MacNab in Proc. Irish Acad. ii. (1877). Engelmann, Revision of the Pinus (1880), p. 165. Coulter and Rose in Botanical Gazette, vol. xi. (1886), p. 256. Stahl, “ Ueber den Einfluss d. Beleuchtung," in Sitzungsb. Jenaisch. Gesellsch. f. Med. und Naturwiss. (1882). Karlston, * Das Traosfusionsgewebe bei den Coniferen,” Bot. Centralblatt, 1879, p. 730. Dufour, “Influence de la lumière sur les feuilles," in Ann. Sc. Nat. (1887), p. 314. Daguillon, “ Obs. sur la structure des feuilles de quelques Coniféres," in Bull. Soc. Bot. France (1888), t. xxxv. p. 57.—Picea excelsa, Abies bracteata, Taxus baccata. The author describes the varying structure of the leaves on the leading and on the lateral shoots respectively. dem, “ Recherches Morphol. sur les feuilles des Coniféres,” in Revue Gén. Bot. t. ii. April 15, 1890, p. 154. P. Klemm, ** Ueber den Bau der beblütterten Zweige der Cupressineen,” in Pringsheim’s Jahrbuch, xvii. (1886), p. 498, t. xxxvii. A. Mablert, * Beiträge zur Kenntniss der Anatomie der Laubblätter der Co- niferen,” in Botanisch. Centralblatt (1885), p. 54 et seq. E. Henning, Analysis in Journ. R. Microscop. Society (February 1888), p. 78. De Bary, Comp. Anat of the Vegetative Organs (ed. Bower and Scott, 1884), pp. 300, 380. R. v. Wettstein, “ Ueber die Verwerthung anatom. Merkmale zur Erkennung hybrider Pflanzen," Sitz. d. k. Akad. d. Wissensch., Band xcvi. (1887), cum ic. G. Kraus, * Mikroscop. Untersuch. über den Bau lebender und vorweltlicher Nadelhilzer,” in Würzburger natur. Zeitsch. v. (1864) tab. 5. (Not seen.) Geyler, * Ueber das Gefüssbundel Verlauf in d. Laublattregionen der Coni- feren,” in Pringsheim’s Jahrb. vi. (1867). G. Sanio, “Anat. d. gemeiner Kiefer,” in Pringsheim’s Jahrb. ix. (1873), p. 87. Goebel, “Ueber d. J ugendzustande der Pflanzen,” in Flora, 1889; also ' Beitrag zur Morphol. u. Physiol. d. Blatter," in Bot. Zeitung, xxx. 1880. Meyer, Die Harzgánge in Blatt d. ABiet. Königsberg (1883). (Not seen.) Noack, * Der Einfluss des Climas auf d. Cuticularisation und Verholzung der Nadeln," in Pringsheim's Jahrb. wissen. Bot. xviii. 1888. (Not seen.) Coulter, « Histology of the Leaf in Zaxodium,” in Botanical Gazette, March 1889, p. 76, T2 256 . DR. M. T. MASTERS ON THE MORPHOLOGY, spora tetragona aurea (a form of Thuya obtusa), the median and the lateral leaves are equally conduplicate, so that the brauch system is four-cornered instead of flattened as it usually is in the genus Thuya, and which finds a parallel in the case of Lycopodium telragonum. In other forms of Thuya, such as T. plicata and T. Wareana, hort., the lateral leaves (always flattened) are so in- ordinately so that they become almost like the median ones in appearance. In Sequoia gigantea, Taxodium distichum, and Fitz- roya patagonica the leaves are all of one pattern throughout. On the other hand, in the so-called genus Retinospora, comprising species belonging to the genera Thuya, Chamecyparis, and Juni- perus, the leaves may occur in three or more different forms. In- dividual plants propagated by grafting or cuttings may present one type of foliage only, and thus it is that different species and even a separate genus have been created on what are only forms peculiar to certain stadia or stages of growth. Two or more forms have been observed on the same plant, or the plant has been watched from the seedling state and the pleiomorphism thus detected. Engelmann * mentions seven descriptions of leaves in Pinus :— 1, cotyledons; 2, primary leaves; 3, bud-scales, which he calls bracts ; 4, the true or secondary leaves ; 5, the scales constituting the sheath of the fascicles of leaves; 6, the bracts of the male inflorescence; and 7, the bracts outside the carpellary scales. These seven may, however, all be reduced to two categories, leaves and leaf-scales. There is a frequent correlation between the form of the branch aud that of the leaves; thus, in JLibocedrus tetragona the branehes are subterete and the leaves uniform in shape, spreading and regularly arranged in four rows. Similar regu- larity of form and disposition occurs in Taxodium distichum var. imbricaria, in Thuya filifera, hort., Abies Pinsapo, Ze, On the other hand, where the branches are flattened, either from side to side or in the median plane, the leaves are generally appressed and unequal in size, the median ones being smaller and flatter than the lateral ones, which are often conduplicate. In Libocedrus Doniana the sterile branches are flattened and bear appressed dimorphic leaves ; the fertile branches bear spreading leaves only, and these pass by imperceptible gradations into the scales of the * Engelmann, Revision of the Genus Pinus, p. 163. ANATOMY, AND LIFE-HISTORY OF THE CONIFER. 257 cone. Illustrations of these facts, as well as numerous transitional stages, may be seen in various species of the genus Thuya, in Li- bocedrus decurrens and L. macrolepis. These homomorphie leaves may be compared with the primordial leaves produced out of season, as will be more fully alluded to in the following para- graph. They may also be compared with the similarly uniform leaves of the species of Lycopodium. Primordial or Protomorphic Leaves.—In many cases the leaves which immediately follow the cotyledons differ in form, attach- ment, arrangement, and, to some extent, even in structure from those which characterize the adult state of the trunk or branches. In Callitris, Frenela, Libocedrus, Thuya gigantea, Cupressus glauca, C. sempervirens, C. nutkaensis, Lawsoni, funebris, macro- carpa, Cryptomeria japonica, Sequoia gigantea, Phyllocladus, Araucaria excelsa (see fig. 3, p. 234), some species of Abies, as in A. Apollinis and grandis, in Pseudotsuga Douglasii, Picea excelsa, Menziesii, orientalis, Larix Griffithii, Cedrus atlantica, the pri- mordial leaves are in many rows, more or less linear, and entirely free or but slightly concrescent at the base, but never appressed. In the Piceas the pulvini are not observable at the base of the lowermost leaves. The stomata are usually on the lower surface of these primor- dial leaves, but sometimes on both, as in many species of Pinus*. In Cedrus and Larix the primordial leaves are linear or awl- shaped, mucronate, with longitudinal rows of stomata on all sides, as in Picea. In Taxus baccata, Picea Menziesii, Tsuga Mertensiana, Sequoia gigantea, Abies Pinsapo, &c. the primordial leaves do not greatly differ in appearance from the adult leaves. In some species of Araucaria, as previously mentioned, the * Daguillon, in a paper “Sur le polymorphisme foliaire des Abietinée,” Comptes Rendus, 1889, p. 108, Jan. 14, published since this paper was presented to the Society in November 1888, says the stomata are always to be found on both sides of the primordial leaves, even in P. Strobus, where in the adult con- dition they occur on the upper surface only. In Abies, in which genus the adult leaves are in one plane, the same author describes the primordial leaves as forming a whorl, whereas in Picea they are alternate, as are the adult leaves, as also in Cedrus and Larix, where the adult leaves are tufted. "The absence of hypoderm and the generally diminished or arrested differentiation of struc- ture as compared with the permanent leaves are also alluded to by M. Daguillon. See also Beissner, * Ueber J ugendformen v. Pflanzen speciell von Coniferen, In Berichte d, deutsch. Bot, Gesellsch, vi. 1888, pl. xxxiii. D to 58 DR. M. T. MASTERS ON THE MORPHOLOGY, primordial leaves are different in character from the adult ones while in other species of the same genus, as in A. imbricata and A. brasiliensis, the form of the primordial leaves is practically the same as that of the adult plant (see anté, fig. 3, p. 234). In Ginkgo the primordial leaves are tristichous, alternate, remote, and pass gradually into those of the adult stage (see ante, fig. 1, p. 233). . In Sciadopitys the primordial leaves are alternate, spreading, long, linear-oblong, and in subsequent stages become reduced to ovate convex scales differing only from perule in being rather more leafy in character. In Pinus the primordial leaves * were recognized by Linneus in Pinus Pinea (see fig. 8). They are, so far as known in various species, flat, linear-lanceolate, generally serrulate at the edge (almost exactly as in the leaves of Lycopodium serrulatum), and arranged as in Sciadopitys. In some species, as in P. rigida, P. silvestris, as also in the Douglas fir and doubtless in many others, as a result of injury to the main shoot, the axis commences to branch at the very base just above the cotyledons, and these primary branches produce only primordial leaves. When a young shoot branches at the base, so that there are belonging to the same generation a central shoot and a whorl of lateral offsets from it, then it often happens, as in Pinus insignis, that in the quicker-growing and more vigorous central shoot the fasciculate leaves are at once produced, while the side-shoots of weaker growth bear primordial leaves only, without buds or fascicles in their axils. Usually it happens that the primordial leaves gradually pass into the position of perula-like scales, or, as they have been called, squame fulcrantes, when they have in their axils tufts or fascicles of leaves. Occasionally these leaf-scales, by an excess of deve- lopment, develop into true linear leaves arranged spirally, and each separated from its fellow by an internode of considerable length. Leaves of this character occur frequently on the lower part of the shoots of the year, as in Pinus Sabiniana, Pinea, sil- vestris (sometimes), and other species. In some cases even they are produced from old woody branches or even from the trunk, as in Pinus edulis, Parryana, rigida, khasyana. * See Linnzus, Syst. ed. Gmelin, 1791, p. 1072; Tristan in Ann. du Muséum (Mém. sur le genre Pinus, p. 242); Engelmann, Trans. Acad. St. Louis (1880). vol. iv. tab. i., Pinus Elliotti. 260 DR. M. T. MASTERS ON THE MORPHOLOGY, Mirbel, speaking of the buds of Pinus, says :—“ Ces bourgeons naissent dans l'aiselle des véritables feuilles, lesquelles sont sèches minces et fugaces dans les arbres qui ont atteint leur troisième ou quatriéme année, mais sont vertes et herbacées dans ceux qui n'ont qu'un ou deux ans . .. les véritables feuilles . . . dans l'aiselle desquelles naissent les bourgeons . . . . acquiérent quelquefois sur les branches des vieux pins la forme et la consistance qu'elles offrent toujours sur les jeunes tiges; et c'est ainsi que les espèces du genre reprennent, comme par accident, les caractéres qui semblent étre les plus conformes à leur organisation primitive." Richard also, in his “ Mémoire sur les Coniféres,” says, ‘ Les feuilles sont d'abord solitaires dans les Pins, pendant la premiére et la seconde année aprés leur germination," alluding, of course, to the primordial leaves *. A similar formation of leaves homologous with the primordial leaves is seen, in some cases, on the branckes or stalks immediately supporting the cones, as in Pinus excelsa &c. The structure of the primordial leaves of Pinus, although essen- tially the same as that of the secondary leaves to be hereafter noted, yet presents some interesting features. The epidermis is papular aud there is no hypoderm. The mesophyll is made up of spheroid cells, the walls of which are not infolded like those of the second- ary leaves. The endoderm and enclosed pericycle are not well marked off, but the central fibro-vascular bundle is perfectly con- formed and has the same essential structure as in the secondary leaves, with this important exception that, whereas in the secondary leaves the bundle generally divides into two, in the primordial leaves it remains undivided. The resin-canals, when present, are sometimes placed in a different position to that which they occupy in the secondary leaves; thus, in Pinus Lambertiana the resin-canals of the primordial leaves are close to the lower epidermis, while in the more permanent leaves they are in the centre of the mesophyll. The mechanical structure as above in- dicated is less highly developed than in the secondary leaves, which have to bear a greater strain. * Mirbel," Observ. sur la famille des Végétaux, Coniféres,” in Ann. du Muséum d'Hist, Nat. xv. 1810, p. 475. Henry, “ Beitr. z. Kenntniss d. Laubknospen," in Act. Acad. Leopold.-Cur. xix. P. i. p. 93, tab. xii. fig. 6 (1837). M. Kronfeld, “ Ueber Polyphyllie b. Pinus Mughus u. silvestris,” in Verhandl. d. k. k. zool.-bot. Gesellsch. in Wien, 38, Bd. iv. 1888. (Not seen.) -——— — ANATOMY, AND LIFE-HISTORY OF THE CONIFERE. 261 Sometimes leaves of a similar character to the true primordial leaves occur on the adult plant, either universally, as in some of the Retinosporas before alluded to, or in association with leaves of the ordinary appearance, when the foliage becomes in conse- quence dimorphic. Thus, in Juniperus sinensis, Cupressus macro- carpa, Retinospora leptoclada (hort.), &c., it is very common to find on some shoots small appressed leaves and others spreading, relatively long and awl-shaped*. The two forms of leaves sometimes occur on the same branchlet, and the degree of con- crescence is relatively small. A similar thing occurs in some species of Frenelat. In these cases the long awl-shaped leaves are of the same form and have the general appearance of the primordial or protomorphic leaves, as is clearly seen in seedling plants of Juniperus pheenicea, in which in the adult state the four-cornered branches are covered with densely crowded, ap- pressed, 4-ranked leaves, while the seedling plant has linear spreading leaves. In some species of Dacrydium, as in D. elatum, Colensoi, and Kirkii, there are remarkable differences between the young leaves, which are spreading and more or less elongate, and the mature foliage, which is short and appressed f. Kirk, in his monograph of Dacrydium §, makes two divisions of this genus. In the one, the young leaves are terete and spreading, and pass by gradual transitions into the mature imbricating state, while in the other the young leaves are linear, flat, and pass abruptly into those characteristic of the adult condition. The genus Podocarpus presents illustrations of similar diversity ; thus, in P. cupressina and P. dacrydioides, &c.||, the leaves on the younger branches are spreading, broadly linear and falcate, while in the more mature state they are polystichous, small and appressed. * It is interesting to notice that Dioscorides, lib. 1, cap. 88, recognized the two forms :—“ Sabina duorum generum est, una foliis cupresso similis spinis horridior graviter olens acris et fervens arbor est coactæ brevitatis que sese magis in latitudinem fundit...... altera tamarisci folio similis est.” See also Chabreeus, Stirpium Icon. p. 72, where both forms are figured. t See Pasquale, loc. ante cit, and Brongniart et Gris in Bull. Soc. Bot. France, xvi. p. 327, 1 Hooker, Icones Plantarum, tab. 1219. $ Kirk in Transact. N.-Zealand Institute, xi. (1877) p. 383; also in his * Forest Flora of New Zealand ’ (1889), with many illustrations, ! Brown and Bennett, Plant. Javæ, t. 10. 262 DR. M. T. MASTERS ON THE MORPHOLOGY, Araucaria Cookii presents an illustration of trimorphic foliage, the primordial leaves immediately following the cotyledons are concrescent, widely spreading and linear; those on the main stem or leader-shoots are oblong acute, and much longer than those on the side branches, which are ovate, acute, densely packed, and appressed. In the ordinary torm of Cryptomeria japonica the concrescence of the base of the leaf is very marked ; but there are varicties in cultivation analogous to the so-called Retinosporas, in which the foliage over the whole tree is of a different character to that of the type, as in C. elegans, C. Lobbii, and C. torta of gardens. The cones of C..elegans, a form analogous to Retinospora squar- rosa, do not materially differ from those of the type, nor does the anatomical construction of the leaf (apart from difference in the shape) differ from that of the normal type. The internal structure of the two forms of leaves is, indeed, practically the same, except that the stomata are usually much ‘more abundant on the spreading than on the appressed leaves. The stomata often occupy a relatively large triangular area at the base of the spreading leaves. This circumstance leads to the conjecture that one purpose of the presence of these spreading leaves is to secure a freer access of light and a fuller elimination of gases through the stomata, and that they may in some way act, as in facilitating the storage of nutritive matter of which the quickly growing extension shoots may avail themselves (see anf, p. 249). Relation of Protomorphic Leaves to Congenital or to External Conditions. The appearance of these protomorphic or primordial leaves in the adult condition and their persistence in Retinosporas can only be attributed to an arrest of development, or to the absolute non-occurrence of developmental changes. It is interesting to correlate these forms of leaves with the relatively slow growth of the branches and the disposition of the stomata, the different position and grouping of which as compared to that in the ordinary leaves, points to some functional diversity. The relation between the primordial and less highly specialized leaves and those which characterized their progenitors is an interesting subject for enquiry, but one upon which as yet little but conjecture can be hazarded. The superficial resemblance of the primordial leaves to those ANATOMY, AND LIFE-HISTORY OF THE CONIFERÆ. 268 of Lycopodium and Isoetes is obvious. In fossil plants referred to Libocedrus, Glyptostrobus, Callitris, Cyparissidium, and other genera from geological formations of different epochs, the same pleiomorphism in the foliage may be observed as in existing genera. It is therefore probable that these diversities in foliage are not so much the result of direct inheritance from an ancestral condition, or of a sudden reversion to it, as of some circumstances which at one time arrest, at another stimulate development and growth. If this be so, the primordial or protomorphic leaves are the outcome of a similarity in the euvironment, common alike to archaic and to existing time. This is borne out by the fact that where growth and development are both very rapid, as in those precocious buds which develop into shoots during their first season of growth, or those which result from injury, such as the removal of the terminal bud by the pruning-knife or otherwise, there the bud-scales exist in the form of primordial leaves. Leaves on the Fertile Branches. Apart from the occasional transitions to be met with between the ordinary foliage-leaves and those of the male or female flowers respectively, perfectly formed leaves may be found on the cone-bearing shoots of a different form to those on the barren shoots. This happens in some silver firs, as Abies amabilis and A. subalpina. In A. firma the leaves on the fertile branches are different both in form and internal construction from those on the sterile one; thus, in the form of the Japanese A. firma known as bifida the leaves are longer and deeply notched and the resin- canals are subepidermal, whilst in the typical firma the leaves are shorter, blunter, and the resin-canals are in the centre of the mesophyll. As each form of leaf occurs (for a time at least) on Particular trees to the exclusion of the other, it is no wonder that they were considered to characterize two different species, A, firma and A. bifida; but cones have never been observed on the latter. On the other hand, the late Mr. John Veitch, Mr. Maries, and others have settled the question by the discovery, in Japan, of the two kinds of leaves growing on the same branch. Indeed, on the leader-shoots of trees growing in this country, leaves of the two forms may be seen on the same shoot, bifid ones below, entire ones above *. * Masters in Journ. Linn. Soc., Botany, vol. xviii. p. 514, where references to the literature of the subject are given. 264 DR. M. T. MASTERS ON THE MORPHOLOGY, Sometimes leaves of a primordial character occur even on the - fruiting-branches ; thus, in Juniperus conferta or J. taxifolia it is common to see fruits borne on branches covered with linear leaves, and I have latterly seen a bush of Retinospora squarrosa (a form of Thuya pisifera with primordial foliage only) covered with cones. Concrescence or Inseparation. One of the most common causes of heteromorphy in the foliage of Conifers arises from the so-called “ adnation " or “ decurrence "` of the leaves. These terms, however, as has been already mentioned under the head of the Cotyledons, are misleading, inasmueh as they imply a previous isolation and subseqnent union. The apparent union is the result of a continued con- nection between the base of the leaf and the branch from which it springs, instead of a separation or detachment such as usually takes place. It is, in reality, due to the arrest or non- occurrence of developmental changes, coupled often with an enhanced degree of mere growth. The leaf is an outgrowth from the super- ficial part of the axis, consisting in its fully developed state of cellular tissue, encompassed by epidermis and traversed by fibro-vascular tissue. The cellular and fibrous elements either become detached from the axis at about the same level or at considerably different levels. The term concrescence, as used by Van Tieghem, is more in accordance with the actual facts of the case. The degree of concrescence varies greatly, the species of Frenela and Callitris being especially remarkable for the extreme degree to which the apparent union is carried *. Among the Lycopodiacew, L. casuarinoides is noteworthy for the resemblance in the arrangement of its leaves to that of Callitris. In the “pulvinus” of Picea, the cellular portion, though not fully separated, becomes prominent at some distance beneath the point where the vascular cords leave the axial cylinder; hence a section of the pulvinus shows it to be cellular for nearly all its length and to be rather a production from the cortex than an essential part of the leaff. The pulvini there- fore require to be distinguished from ordinary “ decurrent ” i * See Parlatore, Studi Organografice .... delle Conifere (1864), tab. 3. . 45. t Masters in Journ, Linn, Soc., Botany, vol. xvii, p. 547 (1879). ANATOMY, AND LIFE-HISTORY OF THE CONIFER. 265 or concrescent leaves, and which have nearly the same structure as the free leaves. Thus a cross section of the concrescent appressed leaf of a Juniper shows the dorsal or outer surface to be convex with a central depression. The epidermis is without stomata, next to it is a layer of hypoderm-cells, followed by palisade-cells filled with chlorophyll, and these by loose parenchyma, through the centre of which runs the fibro-vascular bundle. Near the upper or inner, more or less appressed surface chlorophyll-cells again appear, but less regular in shape and dimensions than those beneath. There is no hypoderm, and stomata are wanting except in a central spot marked by glaucous bloom at the base of the leaf. In these leaves, then, the palisade-cells are on the dorsal surface, which, owing to the position of the leaf, is most exposed to the light, and the stomata are on the upper surface. The relatively free leaves have essentially the same structure. If the bud, say of a Juniper or a Cryptomeria, be examined in the young state, the three leaves of which it consists will be found free at the base. Hitherto growth and developmental change in axis and leaf have been uniform and proportionate ; but after a short time the basal portions are uplifted with the stem in its upward growth, so that the term “ decurrent" really conveys an idea the exact reverse of the truth. It is moreover obvious that, in Juniperus, Thuya, Dacrydium, &c., the concrescent leaves are more especially (but not quite exclusively) found on the more rapidly growing shoots, those which may be conveniently called, in gardener’s phraseology, “extension shoots,” to distinguish them from the more slow- growing framework shoots, and on which the leaves are free at the base owing to the more uniform and regular progress of development. In a young plant of Frenela, preserved in the herbarium of the British Museum, the cauline leaves are all free, acicular, and spreading, but when the stem commences to branch the leaves are seen to be concrescent. Meehan * considers that this concrescence, or adnation as he terms it, is specially charac- teristic of vigour, while free leaves indicate a state of weakness and arrested growth ; and in a sense this is true, for it is gene- rally on the rapidly growing extension shoots that the concrescent * Proc. Acad. Nat. Sc. Philadelphia, 1868, p. 181; 1872, p. 33; and Proc. Amer. Soc, Adv. Science (1868), Chicago; see also Pasquale, ‘ Della eterophyllia nel Cupressus funebris Napoli, 1872. 216 DR. M. T. MASTERS ON THE MORPHOLOGY, leaves occur. But if the distinction between growth and development be kept in mind, it would seem that the concres- cence is an indication of arrested and irregular development associated with disproportionate rapidity of growth. In the free leaves the balance between growth and development is preserved, the base of the leaf is symmetrical and the parts are all in regular proportion. It is worthy of incidental mention, with reference to the possible genealogy of Conifers, that some species of Lycopodium, as L. annotinum, and Chamecyparis have concrescent leaves and projecting pulvini. In some cases all or a large proportion of the adult leaves are concrescent, as in Cupressus nutkaensis, Callitris, Frenela, &c. In other instances the concrescence is much more apparent in the lateral than in the median leaves, e. g. in Libocedrus chilensis, L. austrocaledonicus, Cupressus Lawsoniana, many Retinosporas, &c., where the branch system is flattened, and where, as already mentioned, the leaves are regularly arranged in decussate pairs. Here there appears to be a relative excess or greater rapidity of growth in the lateral pairs of leaves, alternating regularly with a diminished intensity of growth in the median pairs of leaves. In many of these cases, where the leaves are in decussate pairs the free tips of the lateralleaves are in close proximity to the corresponding tips of the median leaves. This proximity is misleading ; thus, in Thuya gigantea, Libocedrus decurrens, or any similar plant, if we fix upon a median leaf, or a pair of median leaves, as a starting point, it will be found on examination that the leaves next in order of time or origin to those from which we start are not those which are nearest, but those which are removed to some distance by the uplifting process, and conversely that the pairs of leaves nearest to those taken as the starting point are further removed in sequence of production. In other words, the closest fellowship is between any given pair of leaves and the lateral pair above them, not between those which happen to be almost on aline with them. On the long fast-growing leader or exteusion shoots the degree of concrescence and the rapidity of growth are more nearly equal in the lateral and in the median leaves. Secondary Leaves of Pinus. The “ needles " of Pinus, like those of Sciadopitys (which will be treated of in another section), have been considered by some ANATOMY, AND LIFE-HISTORY OF THE CONIFER. 267 as foliar, by others as axial. Thus Tristan considered the “needles " as abortive branches, a view at one time adopted by Meehan, but since renounced in favour of the view that the fascicle consists of true leaves emerging from the side of an arrested shoot bearing a “ dormant bud at the apex” *. The fascicles of leaves originate within the axil of a bud-scale or of a protomorphie leaf. This position then indicates their bud-like character. If examined in a very early stage of growth, they will be seen to consist of an axis whose growth in length is arrested but which produces at the base two lateral and five, six, or more sheathing-scales of perular nature surrounding a number of tubercles (2-5), which it is easy to trace in different buds from the initial stage to that of the perfectly developed needle or leaf. The perule are arranged in spirals, while the needles are verticillate. The number of leaves in a verticil varies from 2 to 5, rarely one only is produced, but this solitary condition is rather apparent than real f. The form of the leaves varies according to the number in each verticil, being plano- convex where there are two only, triangular in other cases. In tracing the development of the several fascicles of leaves along the whole length of a young shoot, although of course the youngest are nearest the apex, yet it is remarkable how nearly of the same size and stage of development are all the fascicles. Occasionally it happens that the growing point at the apex of the contracted branch, instead of remaining dormant, is prolonged into a shoot with primary leaves and leaf-buds, see fig. 9. The anatomical strueture of the needles of Pinus varies slightly in the different species, and has, as also that of the primordial leaves, been already alluded to in a former page. 1t may suffice here to say that in all essentials the structure is that of the leaves with the xylem part of the bundle directed upwards and inwards. The evidence derived from comparative morphology, including teratology, development, and minute anatomy, is entirely in favour * Tristan, Lc. p. 246; Meehan, on the Leaves of Conifers, Proc. Amer. Assoc. Adv. Science (1868), B, p. 1, also in Proc. Acad. Nat. Sc. Philadelphia, May 14 (1868), p. 122, Journal of Botany, vol. viii. (1870), p. 133, and Bulletin of Torrey Botanical Club, August, 1885, vol. xii. p. 82. t Kronfeld, « Bemerkungen über Coniferen,” quoted in Botan. Centralblatt, n. 3, 1889, p. 66, cites instances of variations in the number of leaves in the fascicles of Pinus, I 268 DR. M. T. MASTERS ON THE MORPHOLOGY, of the view that the “needle ” of Pinus is a true leaf borne upon a shoot whose apical development is usually arrested after the formation of the verticil of leaves *. AA 7 Wiz . igo NG y, N4 4» Coy Fig. 9.—Terminal shoots of Pinus developed between the two leaves of a fascicle. * See Henry “ Beiträge z. Kenntn. d. Laubknospen,” /. c. tab. xii. fig. 34 (1837); Meehan in Torrey Botanical Club Bulletin (1885), p. 82. Masters in Gard. Chron. (1885), fig. 171. Dickson on the development of bifoliar spurs into ordinary buds, Bot. Soc. Edinb. (1885), Feb. 12. Foei E ANATOMY, AND LIFE-HISTORY OF THE CONIFERJE. 269 : : Monophyllary Pines. In one form of Pinus silvestris the leaves are apparently solitary, but this appearance arises from the cohesion of one leaf to i — another, and the attachment is usually so slight that the two Fig. 10.— Pinus monophylla, showing on the right the primordial, and on the left the adult foliage, with sections of free and of coherent leaves. LINN. JOURN.— BOTANY, VOL. XXVII. H 270 DR. M. T. MASTERS ON THE MORPHOLOGY, may readily be separated. In Pinus monophylla or Fremontiana the same cohesion between two leaves occurs; but in this plant some of the leaves are, in the adult condition, really solitary, terete, and with a circular pericycle. An examination of the mode of development shows that there are always two foliar tubercles, only one of which is deve- loped, while the other becomes obliterated *. In other cases, as above stated, both leaves are formed, but remain coherent by their edges so as to appear simple. Fascicled Leaves of Cunninghamia. Bertrand f records an instance in which the leaves of Cun- ninghamia sinensis were fascicled like those of Pinus. A short shoot was developed in the axil of the leaf; the lower leaves of the shoot were scaly and formed a sheath ; the terminal (fascicled) leaves were very narrow and of the same structure as the ordinary leaves, but destitute of resin-canal and also of stomata. Bons AND BRANCHES. Arrangement.—The buds of Conifers do not differ essentially from those of other plants in position or arrangement. The very marked peculiarity of the ramification depends, as will be presently shown, mainly on the alternate development and non- development of the buds. A very common feature in the arrangement of the buds is the development of one apical bud at the end of the shoots, whether terminal or lateral, and of a circlet of lateral buds around it at its base. In the erect leader o o shoots the circlet is complete Quy but in the lateral branches o it usually happens, as seen in the species of Abies and Picea, that the uppermost buds of the circlet, if developed at all, e . remain in a rudimentary condition, thus O.. This is evi- o dently connected with the horizontal position of the branches ; but it is curious to note that the uppermost buds, those most * Engelmann in Wheeler, Report, vi. p. 259, Botany of California, ii. (1880), p. 124; Bertrand in Ann. Sc. Nat. tom. xx. (1874), p. 102, tab. ix. figs. 5, 6. Hooker in Gard. Chron. (1886), July 31, p. 136; Masters in Annals of Botany, vol. ii. (1838), p. 126 (anatomy and development). T Bertrand in Ann. Sc. Nat., Bot. sér. 5, vol. xx. p. 113, t. 11. figs. 4, 5. ANATOMY, AND LIFE-HISTORY OF THE CONIFER X. 271 exposed to the sun, are not.developed but are checked in their growth to the advantage of the lower buds. In Pinus, on the other hand, the lateral buds in the first instance are erect like the terminal bud; but as they grow they assume a horizontal direction, as in the species of Abies, but with this important exception, that they generally turn up at the tips as growth goes on, and thus allow of the access of light to the branches beneath. Hence in Pinus we find the circlet of buds, whether on the terminal or on the lateral shoots, complete aud equally developed.on all sides. The relative absence of lateral buds, except near the ends of the shoots, is also a marked feature in the Abietinez. Morphologically the bud is simply the apex of the shoot, and in which longitudinal growth is temporarily checked. This arrest is frequently accompanied by a corresponding check in the growth and in the development of the leaves, which assume the form of perule or bud-scales, the perulz being dilatations of the petiolar part of the leaf. In the unexpanded bud the perule are free at the base, but as the shoot lengthens they are sometimes cast off, sometimes remain attached to it, in which latter case they are uplifted with the growing shoot. - Intermediate stages may often be found between the perule ahd the primordial leaves, showing the homology of the two as further illustrated in a former paragraph. In those cases where there is arrest of growth unaccompauied by corresponding arrest of development as in the Cupressinex, the buds are scaleless or naked. The bud-scales are arranged spirally aud are frequently com- pacted together by a felted arrangement of the hairs or fringed margins of the scales as in mauy Pines, or by an exudation of resin as in many Firs. These arrangements are evidently adaptations for the protection of the young buds from cold or wet. In some species of Spruce (Picea) additional protection 18 afforded by the arrangement of the leaves near the end of the 8hoot, and Which, instead of spreading laterally, are directed Vertically parallel to the long axis of the shoot and thus close over the buds. The form of the buds and bud-scales affords useful means of discrimination between certain species. Thus the long thin pointed bud of Pseudotsuga Douglasii 1a characteristic of that species, while in various species of Abies it v2 272 DR. M. T. MASTERS ON THE MORPHOLOGY, is conoidal or globular, in Pines cylindric, domical, acuminate, covered with resin or not, and so on. In most species of Abies and Picea the bud-scales are more or less coriaceous and oblong; in Pinus there is a “ mother ” scale succeeded by two lateral scales at the base, followed in spiral order by 5 or 6 or more scales, which present great variations in texture and duration, being long, thin, and membranous in P. Cembra, P. excelsa, &c., prolonged into a long acumen in P. Coulteri, subcoriaceous and entire in Picea polita, &c., or papery and lacerate at the edges, straight or ultimately revolute at the tips, as in Pinus Bungeana and P. monophylla, Ze. As these perule serve a uniform and a temporary purpose only they are less liable to variation and modifieation during growth from the operation of external causes than organs of longer duration and more complex function, and hence from their relative invariability their utility for classificatory purposes is greater than might at first be supposed *. Deperulation.—The manner in which the bud-scales are removed or thrust aside by the growing shoot is also worthy of attention. The variations observed depend of course on the relation between the nature of the scales, the amount of resist- ance they offer, and the degree of vigour and direction of growth in the bud beneath. The “characters” so afforded supply indications of a general tendency rather than of absolute constancy. In some cases the bud-scales are least resistent to the pressure of the growing shoot at the apex of the bud, in which case the shoot makes its way through a ring or tube of scales which persists around the base of the branch for a long period. This method of deperulation may be called tubular. Illustrations of it occur in Abies amabilis, brachyphylla, bifida, Veitchii, homolepis, cephalonica, nobilis, magnifica, Lowiana, concolor, Pinsapo, bracteata, Fraseri, Pseudotsuga Douglasit, Picea ajanensis, polita, rubra, nigra, Engelmanni, Morinda, Menziesii, and most species of Pinus. * The phyllotaxis of the bud-scales is treated of by Henry in his classical paper entitled ‘‘ Beiträge zur Kenntniss der Laubknospen," in Act. Acad. Nat. Cur. xix. p. 1 (1837), c. tab. ; see also Kichler, Entgegnung &c. in Sitzungsb. d. Gesellschaft Naturf. Freunde zu Berlin, June 1882, p. 90. The adapta- tion of the bud-scales to climata] conditions is the subject of a paper by Griiss, * Die Knospen Schuppen d. Coniferen u. deren Anpassung an Standort und Clima," Berlin, 1885, a paper, however, of which I have only seen the title. Menge, * Ueber d. Blattscheide der Nadeln von Pinus silvestris." ANATOMY, AND LIFE-HISTORY OF THE CONIFERJE. 273 In other species the bud-scales are least resistant at the base of the bud, and when this happens the bud-scales are pushed off in the form of a cap. This may be called calyptrate deperula- tion. A prevalent tendency to calyptrate deperulation is Fig. 11.—Tubular deperulation to the left ; calyptrate to the right. observable in Abies sachalinensis, Picea pungens, Engelmanni, obovata, Morinda, and. rubra, rarely if ever in Pinus. But the tendency is inconstant even on the same tree, being dependent, as before said, upon fluctuating conditions. In Picea Engel- manni, while the deperulation of the lateral buds is calyptrate, that of the terminal ones is usually tubular. The order of development of the terminal and lateral buds at the ends of the erect or of the horizontal shoots is worthy of atten- tion. The general but not invariable tendency in the Abietineæ '8 for the side-buds to expand before the central or terminal bud, even when that is larger than the other. This may possibly be partly accounted for by the restriction afforded by the circlet of buds around the end-bud. In some of the Pines where the cone is apparently, though not really, terminal, the central bud does not start into growth and develop into a shoot until the originally erect cone bends downwards ; hence the shoot in question is a season behind the cone in development thou gh formed at the same time. The direction of growth of the young shoot also presents (in the Jirst instance) differences in different species according as the side shoots turn from the horizontal position upwards or down- wards; thus in Abies Nordmanniana, A. homolepis, and many others, the side shoots at first grow more freely on the lower surface (hyponasty),whence it happens that the lateral shoots are curved in 274 DR. M. T. MASTERS ON THE MORPHOLOGY, an upward direction, Ne» leaving the as yet unde- veloped central bud as it were at the bottom of a cup. In other eases the curvature of the young shoot is downwards owing to disproportionately rapid growth on the upper side of the shoot (epinasty), and in consequence of which the tip of the shoot 18 directed downwards, X ^ V . This happens in Abies Veitchii, sachalinensis, Picea Engelmanni, Morinda, excelsa, ajanensis, &c. These remarks only apply to the direction of the shoots in the first instance; during growth the originally upturned shoot be- comes inverted and vice versd. In the Piceas, where the young shoots are at first deflexed, the stomata are not so distinctly localized in position. In any case, whether the side branches be at first curved upwards or in the opposite direction, growth becomes ultimately equalized on tbe two surfaces, and the branches then, after various changes of position, assume a straight and nearly horizontal position. In Pinus the shoots turn upwards almost from the first, and are directed nearly in a straight direction. After a time the lateral shoots become more or less horizontal, except at the tips, whieh become curved upwards. In Pinus also the young shoots present differences which are useful for specific distinction, in colour, degree of hairiness, form, &c., some being cylindrical, others with prominent angles, with intervening furrows, with one or more rows of resin-canals; but the most remarkable difference here to be noted is the presence or absence of leaf-fascicles at the base of the shoot. Where the growth is uniform the whole length of the shoot is covered with leaf-tufts as in Pinus Pinea, resinosa, Laricio, silvestris, pungens, contorta, inops, densiflora, Llaveana, hud- sonica, Cembra, ponderosa, edulis, rigida, Pumilio, &c. ; but where growth is disproportionately rapid near the base there the base of the shoot is destitute of leaves for some distance, as in Pinus Strobus, monticola, Sabiniana, excelsa, Peuke, Jeffreyt, densi- flora, Bungeana, muricata, tuberculata, monophylla, pyrenaica, Parryana, Massoniana, mitis, &c, This character, however, 18 ANATOMY, AND LIFE-HISTORY OF THE CONIFERS. 275 no more absolute than any other, for in P. montana (uncinata), while the central shoot is leafy to the base, the lateral shoots from the same cluster of buds are naked at the base. The shape of the young shoot soon after its emergence from the bud is sometimes the sarne as at a later period, as in Picea, where the shoots retain the cylindric shape they had at first; while at other times the shape of the shoot differs from that which it assumes when the leaves are fully expanded; for instance, in Abies firma and A. homolepis the young shoot as it emerges from the bud is spindle-shaped or cylindrie, with the leaves appressed to the axis on all sides, thus exposing to the light their stomatiferous or lower surfaces. As growth goes on the leaves, though really in many rows, arrange themselves in one horizontal plane and become pseudo-distichous, some of them, according to their place of origin, becoming twisted at the base so as to allow of the exposure of the upper surface to the light, while other leaves differently placed require no such tor- sion. Spurs.—The species of Larix, Cedrus, Pseudolariz, and Ginkgo are remarkable for the production of two kinds of branches, the one long and slender with the leaves distributed at intervals, the other short, thick, with the leaves in tufts at the extremities. The former are the extension or leader shoots in which growth aud development are rapid; the latter are analogous to the similar growths in the Apple, Pear, Laburnum, &c., but are (in the Conifers) not necessarily connected with the development of fruit, although in Pseudolarix and Ginkgo, however, the spurs bear the male flowers. In the Larches and Cedars the “ spurs ” either remain as such or lengthen ultimately into extension-shoots. The leaves on the extension-shoots are generally longer and more g'aucous than those on the spurs, and are stomatiferous on both surfaces. The mode of development of the spurs may readily be traced 1n the Larch or Cedar, and confirms the view that the appear- ances are due to the more vigorous growth of the basal and peripheral parts in comparison with the central and apical portione. Thus if the bud at the end of a shoot be examined in October the apex will be found to be dome-shaped. The young leaves emerge in succession from the base of the dome leaving me Pex naked, so that the development of the leaves is centri- petal. 276 DR. M. T. MASTERS ON THE MORPHOLOGY, If one of the lateral buds be examined at the same time, the axis will be found to form not a dome but a cup from whose margins the leaves protrude, those at the upper edge of the cup being the oldest and corresponding to those at the base of the dome. These lateral buds are those destined to form the tufts of leaves on the spurs. The greatest energy of growth is in the one case at the apex of the growing axis, in the other at the base. Cladodes or Phylloclades. The peculiar flattened, often lobed and branching expansions in the genus Phyllocladus originate in the axils of scale-like peruls or of linear primordial leaves similar to what may be ob- served on the seedling plant in succession to the cotyledons. Their position and anatomical structure leave no doubt as to their true morphological nature as branches of a peculiar character. The leaves are concrescent with the cladode, and their arrange- ment is described by Van Tieghem as distichous, concrescent with each other in each row and with the branch, so as to form with it a single flattened branch toothed at the edges. In the upper part the cladodia are arranged spirally in more than one plane. The fact of a branch originating from the axil of a rudimentary leaf, producing a green leaf or a small number of green leaves and then ceasing to grow, is compared by Van Tieghem to what takes place in Sciadopitys *, wherein the leaves are inseparate or concrescent and form a single blade or needle, which has its dorsal surface (phloem) upwards and its ventral or xylem surface beneath. “ Needles” of Sciadopitys.— Functionally these structures are leaves, morphologically they present greater resemblance to axial structures. They occupy the axils of leaves of the first order, and thus correspond in position with the fascicle of leaves in Pinus or the seed-scale in the cones of Abietineæ. If examined in the bud-stage, each appears as a tubercle notched at the apex and placed in the axil of a perula which is clearly homologous with a leaf. After the formation of the notch at the apex no further growth in that situation occurs, subsequent increase taking place by intercalary additions at the base. The * See Van Tieghem, Traité de Botanique, p. 1321. ANATOMY, AND LIFE-HISTORY OF THE CONIFERE. 277 whole bud is of a globose form ; and if a vertical section be made of it, the axis will be seen to terminate in a dome-shaped process from the sides of which emerge the perular leaves, each with a “needle” in the axil. Strasburger describes these perule as traversed by a single mesial bundle, and this is probably true, although in many cases I have found the structure to be wholly cellular. A vertical section through the perula and the needle shows the two to have a common cellular basis, so that it is not possible to say where the one ends and the other begins. The needle is longitudinally grooved in the centre of both upper and lower surface, the furrow being deepest on the lower surface. It is traversed by a double vascular bundle, and on microscopic examination the structure is seen to consist of a bounding epi- dermis with subjacent hypoderm, palisade-tissue traversed by resin-canals, each surrounded by strengthening-cells and over- lying a ground-tissue consisting of what Von Mohl designated iransfusion-tissue with large isolated stelliform sclerotic cells. In the centre are two bundles each with its own bundle- sheath. The pericycle consists of small, globular, closely-packed cells : the xylem is on the underside (contrary to what happens in a true leaf), its rays running obliquely upwards and outwards from the base. This oblique position of the xylem is a fact of some interest in connection with that interpretation of the nature ol the seed-scale of Abietinew which postulates the presence of two leaves arising from a contracted or undeveloped axis and having a direction like that of the vascular bundle of Sciadopitys. With regard to the two-fold nature of the bundle in Sciadopitys, it may be pointed out that a similar two-fold or two-branched bundle occurs in numerous species of Pinus, Picea, Abies, &c. ; but in these cases there is but one bundle-sheath encircling the two divergent masses of xylem and phloem, while in the seed- scales of Abietines no sheath at all is observable. The situation of the xylem at the lower side of the needle and its divergence from the lower surface ^ F are in marked contrast to the roc position of the same elements in the leaves of Pinus, in which the xylem is beneath the upper surface "Ai D and the two divi- ions converge from the base. The different opinions held with regard to the morphological 278 DR. M. T. MASTERS ON THE MORPHOLOGY, nature of the needle of this plant may be classed under three heads, according to which the needle is (a) foliar or conjointly foliar and axial, (5) wholly axial or cladodian, (c) an enation from a true leaf. In spite of the peculiar position of the liberian and of the vascular elements of the bundle, Von Mchl* considered the needles as leaves and compared them to the seed-scales of the Abietinez, in which the same relative position of the elements occurs. Mohl’s explanation is that the needle consists not of one, but of two leaves belonging to an abortive shoot axillary to the primordial leaf as the seed-scale of the cone of Abietines 18 to the bract. He accounted for the peculiar position of the xylem and phloem by supposing that the union of the two leaves took place by the margins directed towards and adjacent to the main axis, so that the under surface, and consequently the phloem, came to be directed towards the axis and the upper, or xylem, surface in the opposite direction. This may be represented by the annexed diagrams :— Fig. 12. Fig. 13. For explanation, see text. wherein the circles marked 1 represent the main axis, those marked 2 the secondary axis, with the mother leaf L below and the two secondary leaves l, J, one on each side of the secondary axis. In fig. 12 the lateral leaves 7, J are represented in their natural position, and the xylem x also. In fig. 13 the secondary lateral leaves 7, Z are indicated in union by their posterior * Mohl in Botan. Zeitung (1871), p. 101. ANATOMY, AND LIFE-HISTORY OF THE CONIFERS. 279 margins and dragged out of position, with the result of displacing, or apparently displacing, tbe xylem and the phloem. Engelmann *, Strasburger f, and Van Tieghem ¢ also consider that the needle is due to the concrescence of the two leaves which spring from an axillary shoot, the development of which is arrested 80 that no trace of a punctum vegetationis or growing-point can be seen in the adult condition. It may be added that no trace whatever of the supposed axillary shoot can be seen at any period by mieroscopical examination of the buds, nor any evidence of the torsion and displacement of the leaves. Bertrand $ considers the needles to be axes with a bilateral Symmetry and compares them with the bud (seed-scale) de- veloped in the axil of the bract of the cone of Abietinee. Celakovsky (in litt.) also considers the double needle of Scia- dopitys to be the morphological equivalent of the seed-scale of Abietinem. The late Prof. Dickson thought that the needles represented phylloid shoots or cladodes like those of Ruscus rather than leaves ||. “ There is nothing,” he says, “in the developmental evidence to prove that the arrested punctum vegetationis is not at the apex of the organ between the two projecting points, and if it be at the apex then the organ must be regarded as a cladode.” Dickson was confirmed in his opinion by the occasional develop- ment of branches from the notch of the needle, as first observed by Carrière T, the branching being due to the renewal of growth at the apex, where it usually remains arrested. Dickson, like all other observers, points out that the position of the xylem and phloem is the same in the needles of Sciadopitys, the cladode of Ruscus, and the seed-scale of the Abietinem. Bower also de- scribes the cladode of Ruscus as growing in a similar manner to that described in Sciadopitys. Goebel adopts a similar view to that of Dickson, pointing out * Engelmann in Sitzb. Ges. naturforsch. Freund. Berlin (1868), p. 14. t Strasburger, Conif. und Gnetac. (1872), p. 382. t Van Tieghem, Traité de Botanique (1884), p. 1320. $ Bertrand in Ann. Sci. Nat. (1878), t. xii. p. 67. | Dickson, “ Phylloid Shoots of Sciadopitys,” Report Internat. Bot. Congress, London (1866), and Journal of Botany (1866), p. 224; “ Foliage-leaves of Ruscus,” Trans, Bot. Soe, Edinb. xvi. (1885), p. 140, t. ix.-xi. Sce also for Ruscus, Van Tieghem, in Bull. Soc. Bot, France, Feb, 22, 1884. ise Carriére in Gard. Chron. (ex Reyue Horticole), May 2, 1868, and March 1, 4. l 280 DR. M. T. MASTERS ON THE MORPHOLOGY, that the position of the xylem is wholly exceptional, if the needle is to be considered foliar *. Eichler (in litt.) looked upon the double needle of Sciadopitys as an enation from the primary leaf and extended the same inter- pretation to the seed-scale of Abietines, as will be mentioned hereafter. The evidence derivable from comparative morphology, terato- logy, and anatomy is summarized in an article in the * Journal of Botany’ by the present writer T. The suggestion may be made that the needle is a shoot, the distal or anterior portion of which (in relation to the axis) is abortive, just as the leaf is considered by M. Casimir de Candolle to be a branch the proximal or posterior part of which is not developed f. In Sciadopitys, according to this view, the cladode would consist of two branches, or of two shoots, in union side by side as far as the notch, the leafy dilatations on each side of the central groove would be simply flattened axes, not displaced leaves, and the occasional presence of a bud at the tip would be only an accidental occurrence and one which in any case would not invalidate the axial nature of the cladode. The subject must necessarily be referred to again in treating of the seed-scale; but it may here be remarked that all the authors, whatever view they have adopted as to its morphology, consider the double-needle of Sciadopitys as the homologue of the seed-scale of Abietinem. The absence of a bundle-sheath in the seed-scales may, however, be cited as one point of difference. A RAMIFICATION. One of the principal peculiarities of the branches and stems of Conifers arises from the presence or absence of promi- nent swellings or * pulvini” (fig.14). In the species of Picea and of some of the species of Pinus the base of the leaf is, as it is called, “decurrent,” the cortical tissues being thickened and forming peg-like projections (see p. 264). In Abies and in the * Goebel, Enc. d. Naturwissensch. 1883, p. 216. See also Bower in Proceedings of the Linnean Society, March 6, 1884. t Masters, “On the Comparative Morphology of Sciadopitys,” in Journal of Botany, April 1884. 1 Casimir de Candolle, Théorie de la feuille (1868). ANATOMY, AND LIFE-HISTORY OF THE CONIFERS. 281 Strobus and Cembra sections of Pinus the cicatrix left by the leaf is in the form of a shallow circular rim. The mode of branching in the Conifers is often 80 distinct in appearance as to give rise to the impression that there must be some essential difference between the ramification of these and that of other plants. This, however, is by no means the case. It is always monopodial, never sympodial or diehotomous. The variations depend pri- marily upon the development of the buds in partieular situations and upon their non-deve- lopment in others. Development and non- "m development occur in rhythmie alternation py g. 14.— Pulvini of as regards time, and in relatively definite ` stem of Picea. positions as regards space*. The unusual = degree of regularity with which these phenomena do or do not occur brings about a style of ramification characteristic indeed, but still not essentially different in kind from that which occurs in many other plants. In many Conifers the relatively large development of the trunk as compared with that of the branches, the apparently verticillate arrangement of the latter, their spreading direction, and their gradual diminution in length from below upwards, give rise to a tree of markedly pyramidal form. The virtually verticillate ar- Tangement of the primary branches is associated with a bilateral disposition of the branches of lower order and with a like arrange- ment of the leaves. In other cases, where the disproportion between the stem and the main branches is not so great, we have, as in many Pines, Yews, &c., the bush form, and this becomes modified into the flame-shaped form, as in various Thuyas and Cypresses, the columnar form, as in young plants of Libocedrus decurrens, and the globose form, as in varieties of Thuya orientalis, Ze, Fas- tigiate forms like the Irish Yew owe their appearance to the general upward direction of their branches. This upward direc- tion is due to the retention in adult life of the juvenile mode of growth. Correlative changes are observable in the arrangement of the leaves, which do not in these cages become twisted at the * See Sachs's Text-book, ed. Vines (1882), p. 239; Goebel, Outlines, ed. Garnsey and Balfour (1887), p. 320. 282 DR. M. T. MASTERS ON THE MORPHOLOGY, base, but retain their many-ranked arrangement and do not become pseudo-distichous, and in the size of the leaves, which is generally less than usual. The total leaf-surface may indeed not be actually less than in the ordinary form, but it is more broken up and scattered to secure iusolation from various directions, whereas when the leaves are arranged in two groups in one hori- zontal plane, the exposed surface lies in the same plane to fit the circumstances. Schübeler * figures a lofty tree of the Common Spruce in which the lateral branches, instead of being deflexed or horizontal, ascend to form an oblong head in place of the usual conical one. Such firs are known to the German foresters as “ Spitz Firs.” Pendulous forms are brought about by the downward curva- ture of the shoots. Occasionally the branches in one part of the tree are arranged in one direction, while in another the disposition is different. Thus in a tree, a representation of which is given in the ‘ Garten Flora’ for 1887, August 15, the upper branches are bent down against the trunk, while the lower ones spread in the usual manner. It is curious to compare this different arrangement of the branches with the corresponding variation in the direction of the seed-scales of the same cone observed by the late Alexander Braun f. Great differences of appearance result also from variations in the amount of branching. In the now commonly cultivated Arau- caria imbricata the amount of branching in the young state is rela- tively little secondary, the branches being few and far between. In sume exceptional forms of the Common Spruce (Picea excelsa) this relatively slight degree of ramification produces a remarkable appearance. The slender trunk gives off at remote intervals very long simple branches which hang down ; the lower ones, as it were, writhing on the ground like so many serpents, hence the German name of Schlangenfichte. It is interesting to note that in these little-branched trees the leaves are of unusually large size, a8 if to make up for the small leaf-surface due to the relative scarcity of branches. Personally I have only seen cultivated specimens propagated by grafting; but the late Professor Caspary, of Königsberg, obligingly informed me that he had seen at various times several * Schübeler, Pflanzenwelt Norwegens, fig. 32. t Braun in Sitzungsb. Bot. Verein. Brandenburg, June 26, 1874; scales in upper halt. of cone deflexed, those in the lower portion ascending. ANATOMY, AND LIFE-HISTORY OF THE CONIFER. 283 wild specimens which had originated as seedlings in Prussian forests. One of these was estimated in 1873 to be about 38 years old and was 6-7 metres in height and 7 metres through. Neigh- bouring trees of the same species and age had branches of the seventh and eighth degree of subdivision, but the variety in question had only branches of the third, very rarely of the fourth order. A very similar form originated nearly fifty years ago in Mr. Cranston’s Nursery at Hereford. To the gentleman named I am indebted for the information that the plant was first observed among a bed of seedling Spruce Firs and was planted out and propagated by grafting *. The occurrence of this form in a bed of seedlings of the ordinary character shows that recent, external conditions were not instrumental in producing the variation, else all, ora majority, of the seedlings would have been affected in the same way. The majority of seedlings produced from a tree of this description at Lilienfeld are of the ordinary character, and only a few reproduce the habit of the parent. A tree of similar character exists in the nursery of Messrs. Lucombe, Pince, & Co., of Exeter, and has produced cones one of which was kindly sent to me. It differed in no wise from the common form f. * Mr. Heale informed me that in 1887 the tree in question at Hereford measured 23 feet 6 inches in height, and about 15 feet in diameter. * The following references apply to the Snake-fir or similar forms :—Loudon, Arboretum, iv. p. 2295 (Abies (Picea) excelsa monstrosa): Jacques in Ann. Soc. Hort. Paris (1853), vol. xlvii. p. 652 (A. excelsa, var. virgata) : Carriére, Revue Horticole (1854), tom. iii. p. 101 ; Traité Générale, ed. 2 (1867), p. 331 (Picea excelsa denudata): Koch, Dendrologie, ii. (1873), p. 237, wherein numerous references to ancient literature are given : Gordon, Pinetum, ed. 2 (1875), p. 10 (Abies excelsa Cranstoni): Caspary in Schrift. d. phys.-dkon. Gesellsch. z. Königsberg (January 14, 1873), 116 ff. 128 ff. (Picea excelsa, Link, var. virgata) : Willkomm, Forsti. Flora, ed. 2, 1886, p. 75. Koch in his‘ Dendrologie, above cited, comprises the vars. virgata and Cranstoni with the Pinus viminalis of Alstreemer (1777), a form with long, slender, pendulous branches, described by Caspary in Schrift. d. phys.-ókon. Ges. zu Königsberg, xix. (1878), t. 5, as Picea excelsa, Link, var, viminalis, Casp. ; see also in ‘ Garten Flora,’ 1887, pp. 469, 552, 1889, pp. 136, 657: Schübeler, Pflanzenwels Norwegens (fig. 27, p. 161), figures under this name a form intermediate between those above-mentioned. C. Wilhelm in Sitzungsb. d. k. k. zool.-bot. Gesellsch. in Wien, Bd. xxxvii. (Feb. 9, 1887); Graf, Fr., Berg, “ Einige Spielarten der Fichte, Schlangen- fichte; astlose Fichte (loosely branched); pyramidale Fichte; Trauerfichte (mournful fir) ; Hiingefichte (pendulous); Kugelfichte (globular); Krummfichte (crooked) or Sumpfichte (swamp-fir); nordische Fichte (P. obovata),” Naturf. Gesellech. f. d. Univ. Dorpat, ii. 8vo, 44 pp.; mit 12 Tafeln, Dorpat, 1887: ez Bot. Centralblatt, Band xxxii, 12, no. 51, 1887. 284 DR. M. T. MASTERS ON THE MORPHOLOGY, An exactly opposite mode of ramification to that just mentioned occurs in other varieties grown in nurseries, such as the Clan- brassil variety. In these cases the trunk and primary branches remain short and comparatively undeveloped, while the smaller branches are greatly multiplied so as to form a low globose bush. These sports frequently originate as “ burrs” or bud-variations from branches of the ordinary character; but when removed they may be propagated by grafting. Similar forms occur in other Conifers, e. g. Pinus silvestris, P. Cembra, P. Strobus, &c., but must not be confounded with the singular outgrowths (Heaenbesen of the Germans) which are the result of the attack of a parasitic fungus (ZEeidium elatinum). Mention has already been made of the leaves on fastigiate shoots and of the curious tendency that some species of Abies have of suddenly producing from the upper surface of their hori- zontal branches shoots which have an ascending direction and in which the leaves are, in consequence, regularly arranged round the axis as in Picea Menziesii, ajanensis, &c. Of like cbaracter are the remarkable cases sometimes seen in forests where numerous branches rise erect from the lateral branches surrounding the main trunk like so many satellites. In these cases tbe original radiating position of the leaves has been lost in course of growth and the leaves are arranged on the satellite branches exactly in the same way as on the parent tree. Abies Pichta at Pampesford produces shoots of this character which are * layered " and after- wards separated as independent trees. Loudon, Goepert, and others figure and describe several cases of this kind in the common Spruce, the Yew, the Arbor Vite, &c., where the horizontal branches have rooted in the soil and sent up secondary trunks of the ordinary character from the branches. The two authors last mentioned allude also to analogous cases of the production of erect stems from the prostrate trunks of firs overthrown by the wind, and I have myself seen numerous erect branches proceeding from a felled trunk of Taxodium sempervirens which had been allowed to lie on the ground after it had been hewn down *. * Loudon, Arboretum (1838), vol. iv. pp. 2298 et seg., figs. 2215-2217 ; Goeppert, Verhandl. d. Beford. d. Garten-Baues in d. k. Preuss. Staaten (1853), i. p. 337, tab. 6. figs. 14-19 ; Scbübeler, Pflanzenwelt Norwegens (1875), p. 164, figs. 28-33. ANATOMY, AND LIFE-HISTORY OF THE CONIFERE. 285 Adventitious buds are commonly found emerging from the trunk below the lowest branches of Larix leptolepis, Glypto- strobus, Pinus rigida, P. muricata, and P. serotina. Were sucha tree felled doubtless these shoots would assume the appearance and mode of growth of main shoots. Sometimes, after injury by the pruning-knife, or other cause by which the terminal bud is destroyed, lateral adventitious buds are formed which assume an erect direction. I have seen this in Pinus Coulteri, P. excelsa ; whilst the peculiar form of silver fir called Abies Regina-Amalie, a form of A, cephalonica, owes its origin probably to a like Cause *, g . Carriére describes and figures an example of this kind in | Arau- caria excelsa, in which the extremity of one of the side-branches had been cut off, with the result that a new shoot was developed near the cut end and which assumed an upward direction. But this tendency towards the obliteration of lateral shoots as wit- nessed in the spake-fir, or the proclivity on the part of lateral shoots to assume the direction and form of erect shoots, is by no means universal among Conifers. On the contrary, it is not un- frequently a matter of Tegret to gardeners and foresters that certain species, such as Abies amabilis, A, bifida, &c., do not de- velop leader-shoots, or, if they do, the shoots die, either because they are overpowered by the superior growth of the lateral branches or from other causes. Tying up a lateral shoot so as to make it assume an erect direction is sometimes, but by no means always, successful. So among the Cupressinex it frequently happens that latera! shoots struck as cuttings, although they grow vertically, yet do Dot assume the form of leading shoots, but retain the two-ranked . distribution of their leaves and their flattened form. But this 18 not without exception f. The arrested development of the terminal buds in Conifers may be Compared to the similar arrest that takes place in the vege- tative cone of Welwitschia. The relatively feeble development of the terminal, as contrasted * Heldreich in Regel, Garten Flora (1860), p.313; Seemann in Gard. Chron. 1861, p. 755; A. Murray in Journ. Roy. Hort. Soc. vol, iii. 1862, p. 144, on the Juonymy of various Conifers. t Goeppert in Act. Acad. Nat. Cur. (1868), p. 34, t. 1. LINN, JOURN.— BOTANY, VOL. XXVII. ` * 286 . DB. M. T. MASTERS ON THE MORPHOLOGY, with that of the lateral buds, is probably the cause of the peculiar beaded or cupped growth of some Conifers, as in Abies subalpina, where the ends cf the shoot each year sometimes swell out into well-marked dilatations, or even into cup-like forms, the terminal growing-point being, as it were, included within, or at the base of, the cup formed by the faster-growing side-buds. A remarkable instauce of this kind occurred in an Araucaria growing in the grounds at Bodorgan, Anglesea, where Í had the opportunity of seeing it. In A. Pindrow and some others the arrest in develop- ment of the terminal bud is a marked feature, and one which interferes with the symmetrical appearance of the tree. In the Cupressinee the branchlets or smaller branch-systems are usually flattened either from side to side, or from above downwards. This appearance arises from the production of the branchlets in one plane only, and from the regular conduplica- tion of the lateral leaves (the median ones remaining flat and appressed). In these plants a certain number of branchlets with their leaves have a sort of individuality of their own, so as to resemble compound leaves, the pinnz being represented by sepa- rate branchlets. This individuality is still further indicated by the fact that groups of branchlets, or branch-systems, fall off singly as the individual leaves of deciduous plants do and by a similar process called in this case “ cladoptosis ” *. Variations further occur arising from the degree of ramifica- tion, as in bi-, tri-, quadri-pinnate ramification. In some cases this pinnate mode of branching may take place regularly on both sides of the branch or on one side only, and in that case generally on the distal side, or that furthest removed from the axis, often (as in Thuya) causing a curvature of the branchlet, whose concavity is directed towards the main axis. A similar one-sided ramification is observable in many Algm, e. g. Plo- camium. l In further illustration of these remarks examples may be cited in various genera. Cupressus Lawsoniana has flattened branch-systems placed in a horizontal, in an ascending or in a descending plane in different varieties. The leaves are in decussate pairs, lateral buds being * Dr. James Stark, “On Shedding of Branches and Leaves of Conifers,” in Trans. Roy. Soc. Edinb. xxvii. p. 651, pl. xliv. ANATOMY, AND LIFE-HISTORY OF THE CONIFERZ. 287 formed rhythmically in some of the leaf-axils, while they are un- developed in others. The sequence of the branches is generally as follows :—Holding a branch in front of one and starting from a median leaf, it will be seen that no branch springs from its axil; indeed, as a rule, the axils of the median leaves, whether anterior or posterior, are empty. The lateral leaf next adjoining on the left-hand side has a branch in its axil, which may therefore be called fertile, while its neighbour on the right-hand side is sterile ; the median leaves next above are, as usual, sterile ; above this the left-lateral axil is sterile, the right fertile, and so on, the shoots being produced at every second node, now on this side, now on that. Variations of course occur, and the arrangement is masked by the nearly equal size of the main and of the side-shoots which produces an appearance of dichotomy, while the shoots appear to proceed, on the same level, from both right and left axils. If care be taken to distinguish between the leaf produced on the main branch and that on the side branch next in order, which at first sight appear to belong to the same axis, this fallacious appearance may soon be detected. Often, as has been already stated, shoots are produced on the side of the branch nearest to the axis, but only on the off or distal side; hence the branch becomes curved, with its concavity turned towards the main stem. Towards the apex of the branch the shoots are given off on both sides. On the long, quick-growing leader-shoots the degree of concrescence of the leaves is equal in the median and in the lateral leaves respectively; while on the more slowly growing branches the concrescence of the base of the lateral is greater than that of the median leaves. In Libocedrus decurrens the groups of branchlets constituting branch-systems are placed vertically, so that the conduplicate leaves, which are really lateral, are placed with their edges looking upwards and downwards, while the flat median leaves are directed laterally. Branching takes place with much regu- larity from the axil of one leaf of every alternate pair, that 18 from the axil of a lateral leaf, first on the left, then on the right, and so on. Thuya gigantea.—This, as seen in English gardens in the young state, is a pyramidal or flame-shaped tree, with the branch-systems flattened horizontally, rarely vertically. The leaves are tetra- stichous and decurrent. The ultimate branches are given off x2 288 DR. M. T. MASTERS ON THE MORPIIOLOGY, from the leaf-axils in this order:—(1) median axil sterile, (2) lateral axil sterile, (3) median sterile, (4) lateral fertile, and so on. At the base of the branches the production of new shoots is for some distance confined to the distal side, giving a curved appearance to the shoot, the convexity being to the further, the concavity to the nearer side of the shoot; but towards the middle, and thence onwards to the tip, the branchlets are given off alternately to the right and to the left. Meehan * points out “that in the most vigorous growths of Thuya gigantea and T. oc- cidentalis a branch appears at the eighth node, and always at the eighth node when the vigour of the branch remains the same. As the axis weakens, the branches appear at the sixth node. . - - With greater weakness, the fourth node gives birth to the branch ; aud, finally, as the plant takes on its frondose flattened form, & branch pushes from every alternate node. But in no case does à brauch push at an odd number ; they are always from the second, fourth, sixth, or eighth node.” If for vigour, rapidity of growth on the leader-shoot be sub- stituted, I concur with Mr. Meehan, as my observations tally with his in the main, though there is no absolute rule in the matter; nor, indeed, does Mr. Meehan assert that there is. __' Biota orientalis has the branch-systems ascending and com- pressed from side to side. The leaves are decussate, and branching takes place from each of the lateral pairs, generally only from one axil, the next succeeding branch being given off on the other side; rarely branches are given off from both sides. - On the vigorous leader-shoots, according to Meehan, branching: takes place from eyery fourth node. esa In Chamecyparis spheroidea, referred. 'by- some authors to Cupressus, and included, with Retinospora,: under Thuya by Bentham and Hooker (Gen. Plant. iii. p. 427, 1880), the branch- ing is peculiar, being umbellate or radiating, the main branch- systems in this case not being in one plane, but partially from the median as well as from the lateral leaves, and thus forming inversely pyramidal or wedge-shaped branch-systems, resembling the wedge-shaped tufts of leaves which occur at the ends of the branches of some of the pines, e. g. Pinus ponderosa.. The penul- timate shoots are branched on both sides and compressed, some in the horizontal, others in the vertical plane.: The leaves are * Meehan, in Proceedings Acad. Nat. Se. Philadelphia, June 25, 1872; ANATOMY, AND LIFE-HISTORY OF THE CONIFERS, 289 uniform and 4-seriate. The branching on the main leader- shoots takes place, according to Meehan, pretty regularly at the fourth node, sometimes from the second, rarely from the fifth, — Thuyopsis dolabrata has very flat branch-systems, the leaves are deeussate, and branching takes place from one or both of the third or fourth lateral pairs of leaves above the one taken as a starting-point. Thuyopsis borealis is, in the young state, a shrub of pyramidal habit, with pendulous flattened branch-systems. The leaves are uniform, in four rows. Branching occurs pretty regularly from one of each of the lateral pair of leaves, now to the right, now to the left, rarely from both sides at the same level. On the main quick-growing shoots, the order of branching, according to Meehan, is the same as in Thuya occidentalis and gigantea. * Sometimes in very stout shoots of this plant," says Meehan, "theleaves will be in whorls of three, and then the branching is on the odd numbers 3, 5, 7, but not in a regular graded series as in its normal condition. I have counted as many as fifteen nodes Without a branch ; and this absence of order in branchiug exists also in J unipers, In these the leaves are mostly in threes, though still decussate, and the branching takes place at the odd numbers and is irregular. Callitris quadrivalvis has four leaves in a whorl, and here again we have the irregular branching of the Junipers. “The result of these observations is that in a large number of cases the frequency or degree of branching is seen to be asso- ciated with declining vigour; that presence of leaves in an Opposite pair is favourable to a regularity of branching on even numbers ; and that whorls of three or more are associated with irregular branching on odd numbers.” Probably the phenomena noted by Mr. Meehan are not so much the result of declining vigour as of alteration in the direction and locality in which the energy is manifested. The long quick-growing extension-shoots have, as their function, the formation of the trunk and principal branches ; the multitude of branchlets is adapted to sustain an even greater number of leaves on whose due action the life of the Whole tree depends. Vigour, or the amount of work done, may e as great in the one case as in the other, though exerted in a different direction. Itis probable that some of the irregularities ‘290 . DR. M. T. MASTERS ON THE MORPHOLOGY, ‘mentioned by Mr. Meehan may be associated with deviations from the usual course of the fibro-vascular bundles in the stem $. . It does not, however, fall within the scope of this paper to do more than incidentally allude to the histology ofthe stem. Goep- pert, Schacht, Bertrand, Renault, and others have investigated the subject, and shown how in certain cases the genera may be dis- tinguished by the number and disposition of the areolar punctua- tions and tracheids as well as by the position of the resin-canals. Dr. Mayr also, in his ‘ Waldungen von Nord-America,’ p. 424, arranges the sections of Pinus according to the construction of the wood. Tae MALE FLOWERS. The male flowers of Conifers consist of a number of stamens, each mostly with a filament and an anther or microsporange. These are arranged in various ways at the ends or on the sides of the shoots, so as to resemble the “catkins” of an amentaceous inflorescence. Much controversy has, indeed, arisen as to whether the seeming amenta of these plants are really to be considered as inflorescences composed of a series of axes bearing numerous monandrous, naked flowers, or as single, polyandrous flowers. Following Linneus, Griffith considered the aggregate of stamens to constitute a single flower. The reasons assigned by Griffith for this opinion were the absence of bracts or scales intermixed with the individual stamens and. the unjointed condition of the filaments. The stamens are serially continuous with the leaves, in spires in the Abietines, or in decussate whorls in the Cupres- sinew. Monstrous examples also show the passage of the stamens to leaves, as in some species of Podocarpus and Araucaria, 38 well as in Abietinex ft. * Van Tieghem, ‘Traité, p. 737, says that the vascular bundles of Tuya form four sympodes which traverse the stem in a wavy or undulated manner. At each node is given off to the left a branch which runs through only one internode before it reaches a leaf. The leaves are thus in alternate pairs; and a transverse section shows six bundles —four cauline and two foliar ; a subdi- vision of one of the foliar bundles would naturally accompany the formation of an additional leaf, as when a dimerous whorl becomes trimerous. t The now generally adopted view that the aggregate of stamens constitutes one flower was, as above stated, held by Linnæus, who included several genera of Conifers under his class Monadelphia (Polyandria), assigning to the male flowers of Pinus a 4-leaved calyx. Griffith, Itin. Notes, p. 376; Mohl, Vermischte ANATOMY, AND LIFE-HISTORY OF THE CONIFERZ. 291 Position of the Flowers.—The male flowers are either distinctly terminal (fig. 15) at the points of the young shoots, as in most of Fig. 15.— Male flowers of Picea orientalis, terminal, brilliant carmine ; anthers crested. Schrift. p. 45 ; Dickson in Trans. Bot. Edinb. vi. (1860), p. 418 ; Eichler in Flor. Brasil. (Conifer), and in ‘ Bluthendiagramme, i. p. 59, and in Engler, Pflanzen- Familien Coniferm ; Strasburger, Coniferm ; Engelmann, Revision of Pinus, p.7 ; Bentham in Bentham and Hooker, Gen. Plant. iii. p. 420; Goebel, Outlines of Classif. and Special Morphol. (1887), p. 323. The Opposite opinion, that the flowers are numerous, naked, monandrous, and arranged in catkins, was held by Zuccarini, Morphology of the Conifers, Ray Society, 1846, p. 48; Parlatore in DC. Prodr. xvi. p. 361; Lindley, Vege- table Kingdom, p. 227; Baillon, Dictionnaire de Potanique, ii. p. 181; ‘A. L. de Jussieu, Gen. Plant. ed. Uster, p. 493. - 292 DR. M. T. MASTERS ON THE MORPHOLOGY, the Cupressinez, or they areborne in the axils of the leaves (fig.16) on the sides of the main shoots, sometimes on those of the year (hornotinous), as in the species of Pinus, at other times on shoots of the second and third years, as in Larix, Pinus, Abies, &c. Fig. 16.— Male flowers of Abies amabilis, solitary, lateral, yellowish ; anthers without a crest, but with a point at the back. There is sometimes a variation in this respect in different species of the same genus, e. g. in Juniperus, and where the axillary leaves are abortive or ill-developed it is not possible in all cases to 88y» ANATOMY, AND LIFE-HISTORY OF THE CONIFERE. 298 without having traced its course from the beginning, whether the inflorescence be terminal or lateral. For practical purposes, the prolongation of the leaf-shoot beyond the flowers is & mark of the lateral position, as its arrest is an indication ofthe ter- minal arrangement. In some cases, as in Ginkgo and other Chinese and Japanese forms, the male flowers are stalked, and arranged in umbellate fashion at the ends of short spurs *. The physiological significance of the position of the male flowers on the shoots of the current year's growth, as in Pinus, or on those of former years, as Larix, as just referred to, has not been worked out ; but it would seem probable that some relation might be traced between this arrangement and the nature of the climate. For instance, ceteris paribus, it would seem probable that the newly expanded flowers on the herbaceous shoots would be more likely to suffer from spring frosts and have a less per- fect store of previously assimilated nutriment to draw on than in the case of flowers produced on shoots of the second year. The flowers produced on the second and third year’s wood would also have the benefit accruing from division of labour, as the her- baceous shoots of the first season's growth would, in such case, be confined to their duty of extending the plant by the length- ening of the branches (extension-shoots). Number and Arrangement.—In some species the flowers are solitary, or dispersed at intervals, as in Araucaria spp., Agathis 8pp., Torreya, Picea, Larix, Cedrus, &c.; whilst in others they are aggregated into close heads, as in Athrotaxis, Sequoia, Tsuga, Sciadopitys ; in long spikes, as in Abies, Picea, Pinus, &c. ; or in single umbels or umbellate cymes, as in Ginkgo, Pseudolarix, Keteleeria, Cunninghamia, &e. Sometimes the flowers are sessile or nearly so, as in Larix; at other times stalked, as in Pseudolarix, Ginkgo, some species of Abies, &c ; and in some species of Podocarpus the floral axis is branched. The following list will suffice to show the most usual arange- ments :— Male flowers terminal, spiciform, solitary :— Most Cupressinee, some species of Juniperus, Taxodium, Sequoia, Athrotaxis, Phyl- locladus (some species), Dacrydium, Pherosphora, Microcachrys, * For Pseudolarix, see Mast. in Journ. Linn. Soc., Bot. vol. xxii. p. 210, tab. x. ; and for Kefeleeria, Pirotta, in Bull. Soe. Hort. Toscan. (1887), p. 274. 294 DR. M. T. MASTERS ON THE MORPHOLOGY, Saxe-Gothea, Podocarpus (some species), Araucaria, Cedrus, Larix, &c. Male flowers capitate :—Taxus, Cephalotaxus. Male flowers lateral, spiciform, solitary or clustered :—Abies, Pseudotsuga, Tsuga, Picea, Pinus, Sciadopitys, Cryptomeria, Tor- reya, Phyllocladus, Saxe-Gothea, Dacrydium, Araucaria sp. Male flowers apparently terminal, umbellate :— Ginkgo, Cun- ninghamia, Pseudolarix, Keteleeria. Male flowers branched :—Some species of Podocarpus, Tsuga (sometimes), Taxodium. Perulation and Phyllotaxy.—In some species there is a direct or abrupt transition between the foliage-leaves and the male sporophylls, as in most of the Cupressines, Frenela, Microca- chrys; &c. When this happens there is, of course, no break in the sequenee of the leaves, the phyllotaxy remains unaltered, foliage-leaves and stamens are alike decussate, and there are no perule or bracts at the base of the flower, which in such cases is sessile or not markedly stipitate. In other species the transition is less abrupt, scale-leaves follow the foliage-leaves, to be succeeded by the true sporophylls. Some- times organs intermediate between leaves and stamens may be found. The formation of perular scales or bracts is, of course, the - result of a temporary arrest of growth. When growth recom- mences, it generally happens that the axis of the flower lengthens, so that the flower becomes markedly stipitate. At the same time it usually happens that the perule and also the stamens are arranged spirally, even though the foliage-leaves may be verticillate: thus in most Cupressines leaves and stamens are alike decussate ; while in other cases the stamens are spirally disposed even where the leaves are decussate, as in Athrotaxis, Libocedrus, Microcachrys, some species of Podocarpus, e. g. P. dacrydioides, and Dacrydium. Engelmann * says that the number of perular scales, or, as he calls them, involucral bracts (calyx of Linneus), * varies in the different species of Pinus from 3-16; but it is fairly constant in the same species. The two exterior lateral bracts are strongly keeled, like those of the sheath of the leaves, and stouter and mostly shorter than the others; the third is placed on the upper side towards the axis of the inflorescence; the fourth on the lower or dorsal side opposite the supporting bract, and so forth. * Engelmann, Genus Pinus, p. 8, ANATOMY, AND LIFE-HISTORY OF THE CONIFERE. 295 The innermost ones not rarely exhibit a transition to the anthers, bearing small or incomplete anther-cells on the lower part of their back. In P. resinosa and P. canariensis find the involucral bracts articulated in the middle.” Engelmann then gives the following numerical table of the bracts: 3—4 in P. silvestris and Pinaster ; 3-6 in P. densiflora; 4 in P. Balfouriana, canariensis, Greggi; 4-5 in P. edulis and Parryana ; 4-6 in P. Pinea and P. halepensis ; 4—10 in P. pyrenaica; 5-6 in P. monophylla ; 6 in P. leiophylla, Laricio, and contorta; 6—7 in P. resinosa, montana, and Massoniana ; 6-8 in P. Strobus, excelsa, Peuke, Cembra, rigida, tuberculata, muricata, pungens, and Banksiana ; 8-10 in P. monti- cola, flexilis, insularis, Chihuahuana, Thunbergii, Laricio, vars. Pyrenaica and austriaca, Coulteri and inops; 8-12 in P. Teda; 9-12 in P. Montezume and mitis; 10 in P. insignis; 10-12 in P. ponderosa ; 10-15 in P. Sabiniana; 12 in P. Merkhusii and Elliottii ; 12-14 in P. khasia, glabra, and australis; 14-16 in P. Lambertiana aud cubensis. Eiehler* gives the following details as to the phyllotaxy of the stamens in various species :—2/5 in Cryptomeria japonica ; 9/8 in Tazus baccata ; 8/21 in Podocarpus Sellowii, Lamberti, Lariz europea; 13/34 in Picea excelsa, P. glauca [?]; 2/7, 2/9, and in 4-5-merous whorls (2/8, 2/10), Pinus pumilio, P. silvestris ; 2/18, 2/15, Pinus nigricans ; 2/27-2/31, Araucaria brasiliana. Form, Size, and Colour.—The variations in these respects pre- sented by the male flowers are serviceable for the discrimination of species and varieties even in cases where they are not of great morphological or physiological significance. Thus in the genus Pinus, for example, the flowers vary considerably in size; they are short, thin, slender or long and thick, straight or spirally coiled, : In colour they are usually of some shade of yellow, from very pale lemon-yellow to deep orange. In other cases they are of a violet or crimson colour. uU As these plants are supposed to be purely anemophilous, it 18 difficult to explain the reason of the marked conspicuousness of the male flowers, unless it be as an attraction to pollen-eating insects, The anthers, or male sporangia, of Conifers are borne either on the sides or on the under surface of the staminal leaf or spo- * Bluthendiagramme, i. p. 60. 296 DR. M. T. MASTERS ON THE MORPHOLOGY, rophyll. This leaf consists of a fila ment expanding above into a connective representing the lamina of an ordinary leaf and which is usually flat, in one plene, with one anther-lobe on each side (fig. 17), or the laminar portion may be more or less expanded at right angles to the stalk, thus forming a peltate expansion as in Tazus, or much more developed at the apical eud than elsewhere, as in most of the Cupressinex. Sometimes the anthers are arranged in a ring around the point where the filament expands into the peltate lamina (Taxus), thus resembling the arrangement of Fig. 17.— the sporangia in Equisetum or Marchantia. In other Anther of eases they are placed at the lower edge of the lamina Sequoia. below the attachment of the stalk or filament, as in Cupressine®, an arrangement which may be compared to that of Zamia. The number, colour, form, and mode of dehiscence of the anther- lobes vary in different genera and species, as does also the ap- pearance of the connective. Some of the principal modifications may here be incidentally mentioned. Number of Pollen-Sacs :—2 in Pinus, Picea, Abies, Podocarpus, Dacrydium, Ginkgo, Phyllocladus, Athrotaxis, Sequoia, Sciado- pitys ; 2-4 in Cupressinee ; 3 in Cephalotazus, Cunninghamia ; 4 in Torreya; 5-8 in Taxus; 8-15 in Agathis; 6-20 in Arau- caria, Ze, Direction of the Anther-lobes:—(1) Parallel with or mostly eontinuous with the filament : Pinus, Abies, Picea, Sequoia, Scia- dopitys, Phyllocladus, Agathis, Pseudolariz.—(2) Divergent at the base, often pendulous from the apex of the filament or pel- tate: Taxus, Ginkgo, Tsuga, Sciadopitys, Cunninghamia, Torreya, Microcachrys, Dacrydium, Podocarpus, Araucaria, most of the Cupressinez, Sequoia, Athrotaxis, Cryptomeria, Taxodium. Dehiscence of the Anthers :—(1) Longitudinal: Pinus, Agathis, Araucaria, Cedrus, Larix, Pseudolarix, Sciadopitys, Cunninghamia, Sequoia, Athrotaxis, Cryptomeria, Taxodium, Callitris, Libocedrus, Cupressus, Juniperus, Phyllocladus, Cephalotaxus, Taxus.—(2) Transverse: Abies (some species), Podocarpus spp., Dacrydium, Microcachrys, Pseudotsuga. The Connective.—The variations presented by the connective are also of value for classificatory purposes. Physiologically the con- ANATOMY, AND LIFE-HISTORY OF THE CONIFERS. 297 nective seems, in many cases, to serve as a protection against wet or cold and perhaps against the intrusion of undesirable insects. On the other hand, when the flowers are expanded the open, brightly-coloured connective may serve as an indication of the presence of food to pollen-eating insects, even although they do not take part in the fertilization of the flower. The following will show the principal variations of which the connective is the subject :— Connective peltate :—Taxus, &c. Connective prolonged, ending in a semi-peltate expansion :— Agathis, Araucaria, Pinus, Cedrus, Larix, Picea, Abies SPP., Sciadopitys, Cunninghamia, Sequoia, Athrotaxis, Cryptomeria, Taxodium, Callitris, and Cupressineæ generally, Microcachrys, Podocarpus, Dacrydium, Phyllocladus, Cephalotaxus. Connective scarcely, if at all, prolonged, or ending in a small knob or linear process :— Ginkgo, some species of Abies (e. g. grandis, Pinsapo), Torreya, Tsuga, Sciadopitys, Pseudotsuga, some species of Pinus (e. g. Strobus). When prolonged the direction of the. prolongation is usually more or less at an angle with the direction of the anther-lobes, and the arrangement is such that in the unexpanded flower the counective of one stamen overlaps the auther next above it, but in Saxe- Gothea, according to the illustration (Lindl. Veg. Kingdom, p. 229 b), the connective is bent directly downwards aud thus Covers over its own anther. Pollen-grains.—The differences in the form of the pollen-grains have been made to serve as points of distinction between the Cupressinee and the Abietinee, being globular in the former and provided with wing-like extensions in the other $. These distinctions, ho wever, cannot be considered as absolute, for Pseu- dotsuga Douglasii, which is certainly Abietineous, has the globular pollen of the Cupressinez. The microscopical appearances of the pollen of various species and the formation of a male prothallium by subdivision of the pollen-cell may here be passed over with mere incidental mention and a reference to the works of Strasburger (Conif. u. Gnetac.). Nor does it fall within the scope of this paper to do more than make passing allusion to the relationship existing between the * Brown and Bennett, Plant, Jave Rarior. p. Kg 298 DR. M. T. MASTERS ON THE MORPHOLOGY, microsporangium of the. Vascular Cryptogams and the anthers of Conifers $. Certain differences may, however, be indicated. Thus in Lyco- podium and Isoétes the sporangia are borne at the base of the upper surface of the leaf, as also in Lepidodendron ; in Selaginella they issue from the stem above the leaf (Goebel); whilst in Conifers and Cycads the anthers occupy the lower surface or lower edge of the staminal phyllome. Goebel even points to an analogy between a special growth of tissue on the under surface of the staminal leaf of Cupressus, Thuya, and some species of Juniperus, and the indusium of Ferns. Relation of the Stamens to the Leaves—From what has been already stated it is obvious that the stamens are strictly homo- logous with the foliage-leaves, their position and arrangement are the same; and although the structure is modified by the formation of sporangia, yet intermediate conditions between the leaves and the stamens are met with as monstrosities, aud similar transitions of a more gradual character are to be met with : thus, in Frenela robusta the uppermost leaves pass insensibly into stamens bearing anthers at their sides and base, and these into true peltate stamens with the anthers on the under surface. A similar sequence is obvious in many species of Juniperus, Cu- pressus, Larix, &c. In Pinus the stamens are serially continuous with the primary leaves and not with the secondary leaves. Fur- ther details as to the homology of the stamens will be found in the section treating of the malformations of the flower. The flowers of Pinus are borne on the shoot of the year and ort- ginate in the axils of spirally disposed perule just as the fascicles of leaves do, so that one group of stamens has the same relative position as one fascicle of leaves. The leaves of the male flower themselves bear the anthers or microsporangia, while in the female flower, as will be seen hereafter, the sporangia are borne, not on the floral leaves (bracts), but on “a something " which originates within their axils. In the Larch the bud-scales are spirally imbricated around the base of a column bearing numerous stamens. These male flowers are placed on the second year's wood and correspond in position * Consult Hofmeister, Higher Cryptogamia, ed. Currey (1862), p. 401, &c. Goebel, “ Entwick. d. Sporangien,” in Bot. Zeit. 1881 ; Outlines of Classification, Ae, ed. Balfour, 1887, P. 325. ANATOMY, AND LIFE-HISTORY OF THE CONIFER. 299 with the “spurs” bearing the tufts of ordinary leaves, and are arranged on the 2 plan *. Tue FEMALE FLOWERS. The essential portions of the female flower consist of one or more ovules which are erect or inverted and composed of one coat investing a central nucellus, and sometimes covered from below upwards by a fleshy tubular or annular aril which grows out from the axis, after the formation of the other parts and during the course of their development (Tuzus Zei What is here called an ovule is considered by Baillon, Dickson, A. Murray, Parlatore, and others to be an ovary destitute of style or stigma. The question of angio- or of gymnospermy is, however, one which I do not propose to discuss in this present paper. Outside the ovule is a scale, the seed- or cone-scale, and outside that again another scale, the bract. Frequently the two scales just mentioned are more or less completely combined so as to give the appearance ofa single organ. The discussion as to the morphological signi- ficance of these several parts may be appropriately deferred until alter a general review has been taken of their appearances in different genera. The simplest cases occur in the Tasen and Podocarpes, as for instance in Pherosphera, in which the bract and seed-scale appear to be one and the same, and are flat, somewhat fleshy, and arranged spirally in spikes at the ends of the branches. Each fer- tile branch here subtends a single erect ovule with no arillus. In Saze-Gothea the structure is equally simple. The flat leaves pass gradually into spirally arranged perule at regular distances apart, for the length of an inch or so below the cone. Above, these perule pass gradually into loosely and spirally imbricated, ovate-lanceolate, fleshy seed-scales with a cavity near the base of the inner surface, from the upper part of which hangs the ovule. E single eoat invests the nucellus without any trace of arillus. The female flowers of Dacrydium Franklinii are almost equally simple, but are arillate. The seed-scales are arranged in spikes and each scale bears au erect ovule invested by a tubular arillus. In Taxus the cone consists of a number of spirally imbricated scales arranged on the 2 plan, and all of which, except one, are * Seo Meehan in Proc, Acad. Nat. So. Philadelphia, July 11, 1871. 3800 . DR. M. T. MASTERS ON THE MORPHOLOGY, sterile. The macrosporange or ovule is erect and borne in the axil of a bract placed so close to the apex of a shoot as to appear terminal, partieularly as the apex of the shoot is arrested iu its development. Eventually the ovule becomes invested by a tubular sheath which grows from below upwards and develops into a fleshy aril, in whose cup all but the apex of the ripe seed is concealed. Hooker * considers this aril to be the outer coat or primine of the ovule, an opinion which the internal anatomy hardly suffices to confirm. The mode of development of the ovule, or, as Baillon considers it, the ovary, of Taxus, as well as of Phyllocladus aud Torreya, has been studied by that botanist t. The female flowers of Torreya consist of an ovule the testa of which becomes fleshy, and at the base of which is sometimes to be seen a short imperfectly developed aril, though in the ripe seed it is not to be seen. In Ginkgo the flower consists of an elongated stalk which bears on either side just beneath the apex a straight ovule or macro- sporangium. The outer covering or test is surrounded by an imperfectly developed aril which remains dry, while the test itself becomes fleshy. According to Van Tieghem the vascular system of the peduncle of this plant has its xylem directed outwards and downwards 1. The resemblance of the flowers of the two last-named genera to those of Cycas is noteworthy. In Cephalotaxus the female flowers are clustered in bracteate heads at the ends of axillary pedicels. The bracts are arranged in decussate pairs, each bract subtending two flowers. The ovule consists of an erect nucellus the coat of which, as well as the base of the bract, becomes fleshy in course of development. Usually only one of the two ovules in the axii of each bract develops. By Bentham this genus is included among the Taxodiez, but its ultimately succulent bract and aril seem to indicate a nearer re- lationship to the Taxes. An approximation to Cycas is also evident; but this is even more marked in the following genus. Phyllocladus.—In this genus the female flowers are borne either * J. D. Hooker in Trans. Linn. Soc. vol. xxii. (1859), p. 138. t Baillon, Recherches Organogéniques sur la fleur femelle des Coniféres, p. 4 (1860). t Van Tieghem, Anat. Comp. de la fleur femelle &c. des Coniféres, &c., &c. pl. xv. fige. 58-62; Ann. So. Nat. sér. 5, t. x. (1869). ANATOMY, AND LIFE-HISTORY OF THE CONIFERÆ. 301 singly, or in clusters on the side of a branched phylloclade which has some external resemblance to the ovuliferous scale of Cycas, which latter, however, is foliar. The phylloclade, being partly fertile, partly sterile, su ggests a resemblance to a similar condition in Osmunda. Each flower of Phyllocladus is axillary to a bract which ultimately becomes fleshy. The flower itself consists of an erect nucellus surrounded by an ovular coat which ultimately becomes hard and bony, and which is itself partially invested at the base by a tubular fleshy aril. In the genus Microcachrys * the decussate leaves pass abruptly into the bracts or seed-scales, but with a change from a decussate to a spiral arrangement. The fruit-scales are ovoid and boat- shaped, becoming ultimately succulent, and each one bears a single inverted pendulous ovule from the incurved upper margin. The seed is partially invested at the base by a tubular, fleshy aril. In the position of the solitary ovule there is a resemblance to Sase- Gothea and to Agathis. Podocarpus is remarkable for the fleshy development of the peduncle and of the bracts that spring from and which are congenitally united with it. The flowers are solitary or in pairs in the axil of one of the uppermost bracts. The ovule is inverted, its funiculus being adherent to the fleshy bract. The structure, according to Van Tieghem (loc. cit. t. 15. figs. 79-86), shows that whilst the bracts have the xylem and phloem arranged as in leaves, » in the raphe of the ovule the arrangement is reversed ?, as is generally the case in the fruit-scales of Conifers. In the Cupressinez and the other tribes which follow, there is never any fleshy aril, and only in one genus, Juniperus, do the seed- scales become fleshy. Moreover, the fruit-scale, which externally Appears in many of the genera to be a single organ, in others Shows traces of a compound nature, a complexity indeed which is revealed in all the genera the microscopic anatomy of which has been studied. The scales of the cone are arranged in decus- sating whorls and in continuous sequence with the ordinary leaves. Sometimes the transition is abrupt without intermediate forms ; at other times there is a gradual passage from leaf to fruit-scale, some of the phyllomes having more of the characteristics of leaves, others of fruit-scales. * J. D. Hooker, Flora Tasmania, tab. 100 a. LINN. JOURN.— BOTANY, VOL. XXVII. Y 302 DR. M. T. MASTERS ON THE MORPHOLOGY, The simplest illustration is that of Diselma, now referred to Fitzroya, and in which the erect ovules are borne in pairs in the axil of each leaf. In a few genera, e. g. Callitris, spp. var., and Actinostrobus, the fruit-scales are verticillate and valvate. The whole cone is to be considered as a single flower, inasmuch as the separate scales are direct productions from the axis, without pert- anthial covering and destitute of articulation. The ripe seed-scales are woody in most of the genera, but ulti- mately succulent in Juniperus. In some cases a longitudinal section shows that the cone-scale is approximately of the same thickness throughout, or at least without any sudden marked change in thickness throughout its length, e. g. Callitris, Thuya, and Biota. In others, as in Fitzroya patagonica, Libocedrus, Thuya sect. Chamecyparis, Thuyopsis dolabrata, the base of the scale is thicker than the apex ; while in Thuyopsis borealis, Thuya (Retinospora) obtusa, and Cupressus the apex (exclusive of any terminal mucro) is thickest. Still more noticeable is the circumstance that while in some instances, Callitris, Thuya, Fitzroya, Cupressus, Juniperus, the fruit-scale is apparently single and undivided, in others, though single below, it is divided above into two superposed lamin often of a different shape, as in Libocedrus tetragona and Retinospora leptoclada, a form of Chamecyparis spheroidea. Anatomical Structure.—lu the scale of Actinostrobus, passing from without inwards, we have a brown epidermis encircling 2 mass of parenchyma containing chlorophyll and traversed by two or more resin-canals. Towards the centre, or rather nearer to the ventral side, is a double row of fibro-vascular bundles. In the lower or outer row the phloem is external and the xylem in- ternal, while in the upper row the phloem is above, the xylem below, an arrangement similar to that in the Abietiner. Cupressus has a similar structure, and indeed the other genera of tue sub- order, as shown by Van Tieghem *. The development of the ovules has been studied by Baillon, who, as before noted, considers them to be ovaries developed in the axil of fertile bracts t. Nature of the Fruit-scale.—ln some of the genera, as we have * Van Tieghem, Traité de Botanique, p. 1327; Anatomie de la fleur des Gymnospermes, p. 275. + Baillon, Recherches Organ. sur la fleur femelle des Coniféres, &c. p. 8. ANATOMY, AND LIFE-HISTORY OF THE CONIFERS. 308 seen, the scale at maturity appears to be single; in other cases there is an appearance which i:ight be supposed to indicate the concrescence of two organs (Libocedrus). Griffith *, who was not aware of the internal structure of the scales, considered the scales of the cone of Cupressus as the equi- valent representatives of the membranous scales [bracts] of Pinus, hence he concluded that they were not carpellary leaves but bracts. This view accords with the uninterrupted sequence of the foliage-leaves and cone-scales, but leaves a difficulty as to the origin of the ovules. Eichler t once held similar views as to the single and foliar nature of the cone-scales in this group. Kramer f considers the fruit-scales of Cupressinee as single independent leaf-organs which in process of time develop out- growths from their upper surface and from which the ovules spring. These outgrowths are therefore regarded as of a placen- tary nature. Van Tieghem, on the other hand, states, from a consideration of the anatomical structure described in a previous paragraph, that the scale is double, consisting of the bract and of the ovuli- ferous scale, which he regards, as in the Abietines, as the first leaf of an axillary bud producing ovules on its dorsal surface aud in union by a sheath of parenchyma with the bract §. l Arcangeli ||, agreeing with Van Tieghem as to the anatomical facts, gives another interpretation of their significance. He con- siders that the two fibro-vascular bundles form part of one and the same organ, i.e. a more or less modified branch bearing female flowers and occasional] y a simple leaf. The tribe Taxodieæ, as defined by. Bentham and Hooker, is an artificial group, scarcely, if at all, to be distinguished from the Araucarieæ except by the erect ovules, but in Athrotaxis the ovules are ultimately pendulous. The fruit-scales are arranged * Griffith, Itin. Notes, p. 376. t Excurs. Morphol., in Flora Brasil. fase. xxxiv. 1863, col. 435. t Kramer, “ Beiträge z. Kenntn. d. Entwick. Geschichte u. d. Anat. Baues d. Fruchtblätter d. Cupressineen u. d. Placenten d. Abietineen,” Flora, November 1885, p. 567, tab. 9, $ Van Tieghem, Anatomie de la fleur des Gymnospermes, p. 215. l Arcangeli, ‘Sur la Structure de la fleur femelle des Conifères, &c.," Con- Dis Internat. Botan, Paris, 1878 (published in 1880), p. 35. > Y D 304 DR. M. T. MASTERS ON THE MORPHOLOGY, spirally, not in verticils as in Cupressinee. The bracts are in continuous spiral sequence with the leaves. The fruit-scales of Cryptomeria are closely united for three fourths of their length with the bract, and are dilated at the end into a roundish, cre- nately-lobed extremity projecting beyond the bract as in Chame- cyparis among Cupressinem. It is noteworthy, that while in the male flowers of Cryptomeria the transition between the leaves and the stamens is abrupt, in the females it is gradual and with some of the phyllomes imperfect or intermediate in character between true leaves and perfect stamens. In Tuzodium, Sequoia, and Athrotaxis * the construction of the female flower is essentially the same as in Cryptomeria, traces of the bract being sometimes quite obliterated in the ripe cone, e. g- Sequoia sempervirens. The Araucariee are scarcely separable from the Taxodiem except in the position of the ovule, inverted in Araucarieæ, erect in Taxodiew. Eichler, in Engler's ‘ Die natürlichen Familien, has, by oversight, included the genus Cunninghamia among the last-mentioned series, but by reason of its inverted ovules it be- longs to the Araucarie:. In the genus Cunninghamia the bracts are arranged spirally, being continuous with the leaves; the seed-scale is relatively small and concealed by the bract. In most Conifers the seed- scales form the most conspicuous part, but in Cunninghamia the bracts form by far the largest proportion of the cone. An examination of the inner surface of the bract in the ripe cone shows the upper free portion of the seed-scale in the form of a thin membranous crest projecting a short distance above the inverted ovules, fig. 18, 8. In an imperfectly developed and unfertilized cone the following appearances were presented :—The lowest bracts were quite like the leaves, but thickened at the base (fig. 18, 1 and 2). Next above these were bracts bearing a small cellular outgrowt of a reddish colour just above the short stalk of the bract. Succeeding scales (fig. 18, 3, 4, 5) had 2-5 similar cushion-like processes, which, as has been said, are purely cellular. It is difficult to be * The resemblance of the cone of Athrotaxis to that of Echinostrobus from the Oolite of Solenhofen is noteworthy ; see Renault, Cours de Bot. Fose. tab. 12. fig. 14. Sequoia is, as is well known, represeuted in the Miocene, loc. cit. tab. 13. ANATOMY, AND LIFE-HISTORY OF THE CONIFERÆ. 805 Fig. 18.— Cunninghamia sinensis, Scales &c., with sections. (Seo text.) 306 DR. M. T. MASTERS ON THE MORPHOLOGY, certain whether these cellular outgrowths apparently proceeding from the bract are rudiments of the fruit-scale or of the ovules. In the ripened cone the fertilized seeds are seen to depend from a transverse membranous outgrowth (fig. 18, 5, 6, 7, 8) pro- jecting from the inner surface of the bract about its centre, and which calls to mind the ligula above the macrosporangium in Zsoéíes. Andrew Murray (‘The Pines and Firs of Japan’ (1863), p. 120), after pointing out that in this genus the bract takes the larger share in building up of the cone, and that it has a “ pe- duncle” (petiole rather) or foot-stalk, continues :—‘ The true scale is to be sought inside the bract ; near the base a transverse ridge will be seen just above the seeds, which, on examination, will be found to be the scale adherent to, but outgrown by, the bract. .... The foot-stalk of the bract appears to belong truly to the bract and not to the real scale, which seems to spring from the bract after it has passed the foot-stalk.” Anatomy of the Bract.—A transverse section through the base of the bract of Cunninghamia shows, going from without inwards, an epidermis with a tuft of simple hairs, then alayer of hypoderm encircling the ground-tissue of closely-packed ovoid cells. Five resin-canals, each surrounded by sclerous or strengthening-cells, traverse the ground-tissue as wellasa few scattered libriform cells. Beyond the centre of the bract is a layer or layers of transversely elongated cells(transfusion-tissueof Mohl). Through this pass two fibro-vascular bundles widely separate from each other, but with no marked endoderm surrounding them. The phloem is directed towards the outer, the xylem towards the inner surface. More cellular tissue follows, then hypoderm, and finally epiderm. Thus the structure is like that of foliar organs in general, though dif- fering in minor detail from that of the leaves of this plant, which have very well marked palisade-cells. The specimens examined by myself were imperfect, as they were taken from unfertilized cones, showed no trace of the fruit-scale proper, and therefore the outgrowths from the bracts were purely cellular, still they were decidedly outgrowths from the bract and not from the axis. Van Tieghem * describes and figures a double fibro-vascular system in the fruit-scales, and in the uppermost part of which the phloem and xylem occupy a position the reverse of that in * Van Tieghem, Anatomie de la fleur des Gymnospermes, p. 301, tab. 15. figs. 77, 78. , ANATOMY, AND LIFE-HISTORY OF THE CONIFERJF. 307 the lower portion, which latter represents therefore a leaf or bract within the same cellular sheath as the fruit-scale proper. In the genus Agathis (syn. Dammara) the condition of things is much as it is in Cunninghamia ; the bracts are in spiral sequence with the leaves, a few intermediate structures being found at the base of the cone. The fruit-scale is not visible externally, the seeds springing, or appearing to spring, direct from the inner surface of the bract. The seed of Agathis is provided with one wing (or sometimes two), and this wing, according to Dickson, is on the right side if the direction of the bracts be dextrorse, to the left if the cone-scales be sinistrorse. This observer has traced the development of the inverted ovule of Dammara from the base of the bract, so that in Agathis we have the same condition as in Callitris. Dickson * considers that there is concrescence between the bract and the peduncle of the flower (ovule). In support of this he shows a double vascular bundle, one division going to the bract, the other to the ovule, but this latter may belong to the seed-scale and be, in fact, its only representative. Eichler, how- ever, figures the bundle which proceeds to the ovule as springing from that proceeding to the bract. Van Tieghem f figures a double vascular system in the bracts of Agathis (Dammara), the "pper series belonging to the seed being completely concealed Within the same epiderm and parenchyma as that which surrounds the vaseular system of the bracts. In Araucaria the bracts are continuous with the leaves, as in Agathis, and are congenitally adherent to, or inseparate from, the fruit-scale, which latter, in the mature cone, largely exceeds the bract, just the opposite of what happens in Agathis. Brongniart and Gris t review the different opinions that bave been held as to the simple or compound nature of tbe scale, and sum up by expressing their assent to the views put forth by Dickson and Van Tieghem as to its double nature, as revealed by the arrangement of the vascular bundles, which is essentially that of other Coniferm, although the bundles for the bract and for the seed-scale remain invested by one undivided sheath of cellular tissue. In Sciadopitys the bracts are iu continuous sequence with the * Dickson, Bot. Soc, Edinb, J uly 11, 1861. * Van Tieghem, Anatomie dela fleur des Gymnospermes, tab. 15. figs. 74-16. ` Bull. Soe. Bot. France, vol. xviii. (1871), p. 185; Dickson, Bot. Soc. Edinb. 1861; Van Tieghem, 4. c. p. 277. 308 DR. M. T. MASTERS ON THE MORPHOLOGY, cataphyllary leaves and are arranged spirally. The young un- fertilized cones show the bract to be lanceolate, thick and fleshy at the base, thin above the middle; the thick portion is persistent throughout, but the thinner portion gradually shrivels and becomes smaller in advancing from the base of the cone upwards: the lowermost bracts have no fruit-scale within them; those imme- diately above the base of the cone produce a transverse line of white, many-celled hairs about the middle of the inner surface and which may be compared to the membranous outgrowth in Cunninghamia. Above these are other bracts having in their azils fleshy semilunar scales, less than half the size of the bracts. At this stage the fruit-scales are entirely cellular and consist of cells arranged in radiating lines. About the middle of the inner surface of the bracts next succeeding, a transverse groove may be seen. In the adjacent scale one ovule is observed, in the next three, in the next 5, 7, and 9 respectively. The number of the ovules and their position is not the same in all the bracts, but if there be only one ovule that one is always in the centre. Structure of the Fruit-scale.—1f transverse and vertical sections be made of the base of the scale of Sciadopitys at this stage, bundles of spiral vessels will be seen corresponding in number to the number of ovules, so that the vessels do not appear till the ovules are about to be, or are already, formed. In the bracts of the upper part of the cone the phenomena are the same, but in inverse order, the ovules disappear gradually from the axils of successive bracts, the vessels are no longer seen, and the fruit-scale once more becomes wholly cellular. In the mature cone the fruit- scale largely surpasses the bract in size, and is confluent with it for nearly its entire length. The arrangement of the xylem and phloem in the bract is the same as in a leaf, and corresponds to that in the true leaves of this species; but in the ripe seed-scale the position is reversed, as in the needles or cladodes (see ante, pp. 253, 276), and as in the fruit-scale throughout all the genera of the order. In the Abietinem the cones are placed on the ends of lateral shoots, and usually show &n abrupt transition between the leaves and the perular scales enveloping the female flower, whilst these scales, in their turn, are serially continuous with the bracts. The nature of the fruit-scale in this group has received a larger share ANATOMY, AND LIFE-HISTORY OF THE CONIFERZ. 300 of attention from morphologists and anatomists than hae, till recently at any rate, been bestowed upon the other groups. Nevertheless, it will be found, on comparison, that the structure of the cone in this group presents no essential differences from that of other groups excepting the Taxacew and Podocarpew. The bract is generally conspicuously separate from the seed-scale inthe young condition. Sometimes it remains very prominent in the adult state, as in Pseudotsuga Douglasii, Abies bracteata, A. nobilis, &c. ; whilst in others it is much smaller than the seed- scale and concealed by it, as in the ripe cones of Pinus, or even disappears entirely. In the other genera of this tribe, Cedrus, Picea, Tsuga, Pseudotsuga, Abies, Larix, Pseudolariz, the bracts are never entirely concrescent with the scale, as in Cupressus, but always more or less separate from the fruit-scale, even in those cases where the latter greatly preponderates. In Pinus the cones do not mature till the second year after their formation ; whilst in other genera maturation takes place during the first season. The cones in this genus are usually lateral; but an appearance of a terminal position results from the young cone growing faster than the terminal bud, and assuming an erect position while the bud is temporarily turned on one side. Usually after a time the cone is deflexed, and the terminal shoot assumes its proper direction ; but in some cases, as in Pinus Lemoniana* and P. cubensis var. tertrocarpa, the bud is arrested in its deve- lopment and the cone remains terminal. Minor differences, available for the discrimination of species, exist, as has been already alluded to. Development of the Scales.—ln the young cone of Pinus "ncinata the bracts are serially continuous with the perulæ or brown bud-scales. These get smaller and smaller till they become mere membranous rims from whose axil (or apparently 80) spring the fruit-scales, each provided with its hook-like ter- mination. In the mature cone in this, as in all other species of Pinus, the fruit-scale forms the larger portion of the cone-scale. The bract is thus formed before the fruit-scale; but the latter speedily overtakes it in course of development. At first the bract and the seed-scale are distinct ; but by intercalary growth at the base there appears to be formed a little stalk bearing the * Sir C. Lemon in Trans. Hort, Soc. vol. i. p. 512, t. 20; Murray in ‘ Pine- tum Britannicum, fase. v. 310 DR. M. T. MASTERS ON THE MORPHOLOGY, two lamins (bract and seed-scale). The course of development in Pinus Strobus is the same. Baillon* gives a similar account of the development in Pinus recurva, as does Oersted t of that of P. pumilio; but what he speaks of as the median lobe of the axillary scale, and which, he says, is developed subse- quently to the other part of the scale, appears to be the true apex of the scale, which is terminated by a muero and dilates laterally into the umbo. The thickened end or apophysis of the cone-scales is peculiar to some sections of the genus Pinus and is an after-development. At first, as in P. Laricio, the scale is nearly flat, as it remains in P. Strobus, P. Cembra, &c. ; but as growth goes on, the activity of growth is most manifested on the under or outer surface, so that the scale becomes unsymmetrical and thicker below the (originally) central umbo than above it. In Pinus Coulteri the scales retain their primitive equality of proportion, and terminate in a central pyramidal point. Andrew Murray regarded this umbo and mucro as the repre- sentative of a petal! The same author also suggested, as an alternative hypothesis, that the bract was the homologue of the petal, the fruit-scale of the disc; while the wing and test of the ovule or seed were regarded by him as pericarp 1. Anatomical Structure.—The anatomy of the bract and seed- scale in the Abietines is essentially the same as in the other sub- orders, the vascular cords of the bract having always the phloem below, the xylem above; while in the fruit-scale the relative posi- tion is reversed, but occasionally, as in Pseudolarix Kaempferi, the bracts retain for a longer or shorter period a purely cellular eondition, in which no differentiation into xvlem and phloem is observable. In this plant the bract and the fruit-scale have a common cellular basis, and the vascular cords pass direct from the axis into the fruit-scale, leaving the bract, at any rate at first, wholly cellular. The essential structure of the fruit-scale, as in all other Conifers, is the same as in the needles of Sciadopitys or the cladodes of Ruscus. In the bract there is generally a single fibro-vascular bundle, whilst in the fruit-scale there are several in one plane. The central axis or core of the cone to which the fruit-scales * Baillon, Recherches Organogéniques, 1860, t. 1. + Oersted, Bidrag til Naaletrernes Morphologi (1564), tab. 2. fig. 35. ł A. Murray in Gard. Chron. (1866), pp. 8, 852. ANATOMY, AND LIFE-HISTORY OF THE CONIFERJA. 311 either remain firmly attached, or from which, as in the Silver Firs (Abies), they become early detached to liberate the seed, presents always the structure of an axis with a central pith sur- rounded by spiral vessels, a hard woody centre of variable thick- ness and consistence traversed by medullary rays, and an outer cortex traversed by resin-canals. MALFORMATIONS OF THE FLOWER. Only a few of these need here be noticed. The plant figured by Trautvetter* as Thuiecarpus juniperinus is interesting as affording an instance of permanent arrest of development. It is in fact a Juniper in which the fruit-scales, instead of becoming confluent and succulent remain free nearly to the base, and do not assume a succulent character. Like Retinospora, it supplies an illustration of a genus founded on characters of a temporary or accidental character. The ordinarily dicecious species, such as Araucaria imbricata, sometimes produce flowers of both sexes on the same treet. Carriére figures and describes a similar occurrence in CepAalotaxus Fortuneit; while in the Yew the monecious condition is not very uncommon. A gradual transi- tion from the leaves to the stamens or sporophylls is also occa- sionally met with, especiall in prolifed flowers of Lariz and other genera$. Braun notes that although female plants only of Taxus tardiva are known in gardens, yet seeds are pro- duced in abundance ||, probably because the flowers are fertilized by the pollen of Taxus baccata. But the most important malformations to be considered, for our present purpose, mE certain monstrosities of the cones, which may be comprised under two heads :—(1) Androgynous cones, and (2) Proliferous cones. Androgynous Cones.—Many instances of this peculiarity are * Trautvetter, Plantarum Imagines (1844), p. 11, t. 6. t Masters in Gard. Chron. (1873), p. 291, c. ic. ł Carriére in Revue Horticole (1878), p. 117. . $ Schleiden, Principles, ed. Lankester, 1849, p. 229. A. Braun, “ Ueber eine Misebildung von Podocarpus chinensis,” in Monatsb. k. Akad. d. Wissensch. Berlin, October 1869 (Podocarpus). (See also Hook. Bot. Mag. t. 4655 ; Oersted in Vidensk. Meddelels. p. 83.) ! Braun in Sitzungsb. bot. Verein. Brandenburg, June 1874, p. 744, adnot. 312 DR. M. T. MASTERS ON THE MORPHOLOGY, upon record *. One of the most remarkable is that of Cupressus Lawsoniana, sent me by Mr. George Syme, a gentleman whose knowledge of living Conifers and their structure is of an extensive character. This specimen was alluded to in my paper on the Morphology of the Primulacew. In it the lower scales o the male flower, which were serially continuous with the leaves, bore anthers; while the upper scales (bracts) of the same flower, also serially continuous with the leaves, bore ovules. One scale (see fig. 19) even bore an anther on the outer and an ovule on the inner or upper surface of its basal portiont. It will be remem- Fig. 19.—Scale of Cupressus Lawsoniana, bearing an anther on the outer and an ovule on the inner surface. Magn. 5 diam. bered that the fruit-scale, such as is met with in the Abietinem, more or less conspicuously, as distinct from the bract, is not appa- rent externally in the Cupressinem, though the internal struc- ture indicates, in all the genera yet examined, the presence ofa double vascular system, one belonging to the bract, the other to the fruit-scale, as already frequently mentioned. In the androgynous specimen now referred to some of the * Mohl, 1837 (repub. in Vermisch. Schrift. p. 45, 1845). Meyen in Wiegmann's Archiv (1838), p. 155. Schleiden, Principles (1849), p. 229. A. Braun, Das Individuum (1853), p. 65. Dickson in Bot. Soc. Edinb., J uly 1860. Caspary, De Abietin. flor. fem. Struct. Morph. (1861). Cramer, Bildungsabweich. (1864), p. 4, t. v. ff. 13-17. Parlatore in Ann. Sc. Nat. (1865), sér. 4, t. xvi. p. 215, t. 13a. .. Oersted in Vidensk. Meddelelser (1868), p. 83. Se Sperk, “ Androgynous Cone of Larix europea,” in Mém. Acad. Imp. 9c. St. Pétereb. t. xiii. (1869), p. 53, t. 1. Masters in Gard. Chron. June 30 (1883). t Masters in Trans, Linn. Soc, ser. 2, vol. i, (1877), t. 41. figs. 9-11. ANATOMY, AND LIFE-HISTORY OF THE CONIFEBR EK. 313 uppermost scales were thickened at the base, the thickened por- tion bearing 2-5 erect ovules; but in other instances this thickened portion was wanting, and the ovule was borne at the end of a little stalk in the axil of the bract, or springing from its base, This was the case with the one scale just mentioned, which bore a single anther-lobe on its dorsal surface and a stalked erect ovule from the base of its inner surface without any trace of the fruit-scale, unless the minute stalk supporting the ovule be regarded as representing it. Unfortunately the minute anatomy and disposition of the vascular system was not examined ; but it is quite certain that the same scale produced anther and stalk supporting the ovule, and that this scale was continuous with the leaves. An interesting comparison may be instituted between this case and the structure of Cunninghamia and of Sciadopitys above referred to. In the instance recorded by Oersted the bracts of the female cone of Picea nigra passed gradually into stamens, so that the cone, which was of the ordinary character in its upper portion, passed below into the condition of a male flower. In the axils of some of these staminoid bracts what Oersted called “ carpels ” were developed, consisting of two squamules of varying form, whilst in the axils of the upper bracts the two squamules had coalesced so as to form a large scale notched at the top, but similar to a carpel. The same author described a cone of Picea alba normal in character at the base, above which the bracts gradually increased in size, while the “ carpels," i. e. the fruit- scales, diminished in size. In a male flower of Pinus rigida examined by myself in 1876 the lower portion was normal, consisting of stamens of the ordi- nary character, but in the middle these were replaced by flat scales. Some of these latter were sterile, while others had in the axils rudimentary fruit-scales, and others again perfect fruit- scales with inverted ovules. In Pinus Thunbergii I have met with similar changes, the base of the female cone occupied with stamens, continuous below with bud-scales and passing above into bracts with a fruit-scale on the inner surface (tig. 20, p. 314). A flower of the common Larch (Larix europea) sent me by Prof. Oliver showed the usual perular scales at the base investing the staminal column, the stamens were more or less phyllodic or bract-like and occupied the lower half of the staminal column, 314 DR. M, T. MASTERS ON THE MORPHOLOGY, the upper portion bearing numerous bracts, the more perfect of which had a central apical lobe prolonged into a long tube or acumen and two smaller side-lobes. Within aud at the base of Fig. 20.— Bisexual cones of Pinus Thunbergii ; with details of structure and ~ pollen-grains. these bracts was a small pedicled fruit-scale of transversely oblong or reniform shape, very much smaller than the bract in every instance, but with rudiments of two inverted ovules. On making sections in various directions ıt seemed, in this case, as if the fruit- scale, though produced in the axil of the bract, was not an enation from it, but an outgrowth from the axis above it, it being mostly quite free from the base of the bract, whilst its vascular system proceeded direct from the axis. Proliferous Cones.—Adverting now to proliferous cones, OF those in which the central axis is prolonged beyond the bracts and scales, it may be said that they are not uncommon, it being rare in some seasons to meet with a tree of Cryptomeria japonica in which some of the cones are not so altered. With the exception of Sciadopitys, to which special attention must be given, the ap- ANATOMY, AND LIFE-HISTORY OF THE CONIFERS. 315 pearances presented by these proliferous cones in various genera are essentially similar. The bracts become more or less leafy, and, indeed, those at either end of the cone pass gradually into the condition of ordinary leaves, so that the general appearance is as a branch growing through a cone. In Tsuga Brunoniana the bracts are completely metamorphosed into leaves, while the scale is reduced to a lenticular process des- titute of-ovules and almost, if not wholly, detached from the bract (fig. 21). The fruit-scales sometimes disappear gradually, at others Fig. 22.—Portion of the cone of a Proliferous Fig.21.— Tsuga Brunoniana, Larch. 2 is a fruit-scale in the axil of an showing “leafy bract ed- unchanged bract; at 3 the leafy bract en- tached from the fruit- closes at 4 a sterile fruit-scale ; at 5 the scale scale, which is barren. En- has disappeared entirely. Enlarged. larged. they are more or less changed, while in still other cases they are not appreciably altered. Thus in a proliferous larch- cone I found the woody scales more or less winged at the sides, notched and bipartite *, and sometimes at the bottom of the notch a short cylindrical column was present, suggestive of the axis of a rudimentary bud, a matter of some significance in relation to the view taken of the nature of the fruit-scale by Caspary and others, to be hereafter alluded to (fig. 23, p. 316). Sometimes * Compare this with the divided fruit-scale of Schizolepis Braunii figured in Renault, Cours de Botan. Fossil. Conif. tab. 12. ff. 1-4. 316 DR. M. T. MASTERS ON THE MORPHOLOGY, the lateral lobes of the scale were infolded so as partially to conceal the ovule at the base and suggest the idea of a partially closed carpel. Vo Be o 4% Fig. 23.—Larir. A, proliferous Larch-cone; s, leafy bract and seed-scale ; €» leafy bract, the scale rudimentary ` D, £, abnormal scales with traces of ovules. Real size, The proliferous cones of Pseudotsuga Douglasii are chiefly re- markable for the fact that in passing into the leafy state the bracts gradually lose the three-lobed apex which usually charac- terizes them. In the herbarium of Linnæus preserved in the rooms of the Society is a specimen of the Common Spruce (Picea ezcelsa) m which the fruit-scale within the bract, and seen from without, 18 represented by a central axis with a lobe on each side. Looked at from the inner side the two side-lobes are seen to be united at the base by a transverse band and to form with the central lobe a cup, occupied by a scaly bud the source of whose origin 18 doubtful. The explanation seems to be that there is a central axis bearing on each side a leaf or leafy process. These processes ANATOMY, AND LIFE-HISTORY OF THE CONIFERZ. 317 are erect, appressed, and somewhat conduplicate or involute, the posterior edge of one united below with the posterior edge of the Fig. 24.— Details from a prolified cone of the Douglas Fir. A, normal bract ; B, the same, seen from the side, attached to the scale at the base; c, bract and seed-scale, from the outside; p, bract from a proliferous cone passing _ into leafy condition ; in x, F the bract is completely leafy, E seen from the outside, r from the inside. other to form a sheath surrounding the bud. They may be con- sidered as leaves or, as seems more probable, as wing-like pro- Jections from the axis (fig. 25). This malformation is similar to those described and figured by Oersted *, by Stenzel, Willkomm, Eichler, and others, in whicb, * Oersted, Z. c. tab. i. figs. 1-15. Stenzel in Nov. Act. Acad. Leopold.-Carol. Band xxxviii. (1876) tab. 4. Willkomm in Nov. Act. Acad. Leop.-Car. Band tli. ii. Th. a2, Halle, 1880, c. tab. Eichler, “ Ueber Bildungsabweichungen bei Fichtenzapfen,” Sitzungsb. d. k. Akad. d. Wissenschaft. z. Berlin, January 1882; and “ Fntgegn ung auf Herrn L. Celakovsky's Kritik &c.," in Sitzungsb. d. LUNN. JOURN.— BOTANY, VOL. XXVII. 3 318 DR. M. T. MASTERS ON THE MORPHOLOGY, with various modifications of detail, a leaf-bud is found within the bract, taking the place of the fruit-scale and possibly in some Fig. 25.—Monstrous flowers of Picea excelsa, in the herbarium of Linnæus. (For explanation see p. 316.) cases developed from it. This latter conclusion, however, is con- tested by Eichler, who considers these buds to be purely adven- Gesellsch. naturforsch. Freunde zu Berlin, June 20, 1882. Celakovsky, Zur Kritik der Ansichten von der F ruchtschuppe, &c. (1882); and also the same author's reply to Eichler,‘ Ueber Herrn A. W. Eichler's Entgegnung,' &c., Prag 1882. Velenovsky, “ Zur Deutung d. Fruchtschuppe der Abietineen," Flora, 1888, p. 516, c. tab., a paper published since this communication was written, considers the normal fruit-scale to consist of two united leaves proceeding from a shoot arrested in its growth. ANATOMY, AND LIFE-HISTORY OF THE CONIFERS. 319 titious productions having no direct relation to any of the normal structures *. Kichler had the opportunity of examining and of comparirg nearly all the cones studied by his predecessors, and gives drawings representing various stages of malformation from a slight notching of the apex of the fruit-scale to its division into three parts, one posterior and two lateral, and the formation of a bud, or even of a shoot, from the internal portion of the posterior lobe. In the instances examined by myself I have not been able to ascertain satisfactorily the precise origin of the adventitious shoot. Eichler's general conclusions are that the fruit-scale is not a complex organ, but an outgrowth from the bract, so that the bract and the fruit-scale are not two different organs, but parts of one; and that the bud or shoot, which in these proliferous cones is often produced in the axil of this bipartite body, is the develop- ment of an axillary bud usually latent. This bud may appro- priately be compared to the buds formed at the summit of the branches in Lycopodium Selago t and on the leaf of Isoëtes. In Lycopodium the sporangium is axillary and erect, often supported on a short stalk. The adventitious buds, above named, are also stalked and flattened in the median plane (as the fruit-scale of the cone is). From the side of the stalk emerge the leaves in decussate pairs, the two lateral leaves of the lowest pair being equal; the anterior leaf of the pair next above is larger than the posterior. The members of the succeeding lateral pair are equal ; of the fourth pair (median) the anterior one is larger than its fellow, and each appears to arise from the base of the cup formed by the succeeding fifth pair. This fifth pair is lateral, and its components are greatly larger than any two that precede or follow them, and they are so twisted at the base that the surfaces, instead of being in the horizontal plane, are turned nearly into a vertical position and directed antero-posteriorly. More- ee * Compare Eichler, Ent 81, figs. i., ii, and p. 92, figs. v.-x. ; > gegnung, l. c. p. 8l, gs. 1, U., i ` also the same author's “ Bildungsabweichungen," l. c. figs. 2-12, and of which fig. 6 almost precisely resembles the condition in the Linnean specimen; figs. 5 and 7 also Present a close resemblance. ‘These figures are taken from spect- mens derived from the Botanic Garden of Upsala, whence p ossibly the Linnean "Pectmen may also have been derived. a oo. t For an account of these buds in Lycopodium see :—Dillenius, Historia Musor, 436; Newman, History of British Ferns (1844), p. 379; Hegel- aer in Botan Zeitun 7 74) p. 513. ` 8g (1872), n. 45, and (1874) p. 9^ . lon The buds of Isoëtes are described by Goebel in ‘ Botanische Zeitung, 1879, n. 1, and in ‘ Outlines of Classification &c.’ ed. Balfour, Oxford (1887), 5 » 295. Z 320 DR. M. T. MASTERS ON THE MORPHOLOGY, over, these two leaves are concrescent at the base so as to form a shallow cup from the outer side of which, a little above its base, emerge, as has been said, the two leaves constituting the pair immediately beneath and which have probably been uplifted by the vigorous basal growth of their larger neighbours. . - The leaves of the sixth pair are median, and this time it is the posterior of the two which is the larger, whilst the smaller anterior one is uplifted with the base of its neighbour and with it forms a shallow cup. The next two leaves are lateral and slightly unequal; the next succeeding, eighth pair is median, and of these the anterior leaf is the larger. Between the third and fourth pairs of leaves the internode is rather more lengthened than between the other pairs, and this allows of the sepa- ration as a '* gemma" of the upper portion of the bud. This description from living specimens differs somewhat from that of Newman, and is introduced here with the view of inducing future obser- vation as to whether there is anything more than superficial resemblance between these buds and the adventitious ones that are formed in some pro- liferous cones. ` LL Goebel describes the formation of buds or shoots in Jsoétes at the position on the leaf where a sporangium is usually found, and this shoot separates from the mother-plant and develops into a new plant, as in the case of the Lycopodium just. mentioned. . In both these instances the formation of adventitious buds seems, like that in the prolified cones, to be due rather to substitution than to any actual metamorphosis of the sporangium, or even of the axis bearing it. Parlatore * describes and figures a cone of Pinus Lemoniana in which the lower scales are changed into branches with leaves, and a cone of Abies Brunoniana affected in a similar manner. In Sequoia gigantea M. Carriére has figured a cone with both leafy and axillary prolification T. The proliferous cones of Sciadopitys are of great interest (figs. 26 and 27). In theordinary cones of this plant the bracts are nearly completely concrescent with the fruit-seale, but in the specimen 10 question the bracts aud the fruit-scale are more or less detached one from the other. Moreover, the bracts gradually assume the con- dition of the perulze such as surround the buds. In this plant, then, the braets, in place of becoming more leafy, as they do usually in proliferous cones, revert to the vaginal or perularcondition. The metamorphosis is in this case retrograde instead of progressive; or, to speak more correctly, development has been arrested instead of enhanced. From the axil of each of these perulz proceeds 4 * Parlatore, Studi Organografice delle Conifere (1864), p. 35. t Carriére in Revue Horticole (1887), p. 509, fig. 103. ANATOMY, AND LIFE-HISTORY OF THE CONLFERS. 321 “needle” or phylloid shoot of the ordinary character, so that in these cones we have it in evidence that the perule are modifica- tions of leaves, that the needles or phylloid shoots are axillary to Fig. 26.— Details from prolifie d cone of Sciadopitys. a, bract and scale from the side and from the back re- spectively; B, section of A, to show d" ; i detachment of scale and bract; Fig. 27.—Proliferous cone of Seiadopitys c, bract; D, bract and portion of (From Veitch’s ‘Manual of Conifers.") "needle? from proliferous cone, seen from the back, he perula that the them, that th the same relation to t , ey occupy the sa d, further, that fruit-scale does to the bract in ordinary cases, an 322 DR. M. T. MASTERS ON THE MORPHOLOGY, they have the same essential structure as the fruit-scale of this and all other genera. The last malformation that needs mention here is one in Pinus muricata, in which a sort of false cone was produced consisting entirely of bracts which were arranged in spirals consecutively with a number of brown membranous scale-leaves on the stalk. The bracts were thick and spongy, green at the base and purplish brown at the tips, and each was traversed by a single fibro-vas- cular bundle proceeding direct from the axis. No trace of fruit- scale or of ovule was to be seen. Whether such a production would have been formed later on is doubtful. Before leaving the subject of malformations it is as well to mention the peculiar cone-like galls made by certain Aphides (Chermes). These are exceedingly common on the Larch and Spruce, and might have been passed without comment but for the fact that a genus was founded on them by one of our most dis- tinguished botanists. OVULES AND SEEDS. Throughout this communication the gymnospermy of the Conifers and their intimate relation to some of the Vaseular Cryptogams has been taken for granted. It is not the writer's intention in this place to enter upon the consideration of these branches of the subject, which is amply discussed, so far as our existing knowledge permits, by Strasburger, Eichler, Baillon, and other botanists frequently mentioned in the course of this paper. The most noteworthy characteristics from our present point of view are the presence of an aril in most Taxacee and Podo- carpe, and its absence in the other suborders. The erect or inverted position of the ovule is also evidently significant, as the whole order may be subdivided into two groups by this eha- racter. The number of the ovules varies, as has been shown. Sometimes it is solitary, asin Zaxus; in Actinostrobus, Libocedrus, some species of Thuya and Chamecyparis, there are two (rarely more) seeds to each scale. In Thuyopsis there are four or five seeds, while in Callitris and Cupressus they are numerous. ln Taxodium there are but two ovules, while in Cryptomeria, Sequoia, and Athrotaxis there are several. In Araucarie, Araucaria and Agathis have solitary ovules, . FEN re ANATOMY, AND LIFE-HISTORY OF THE CONIFKRE. 323 while Cunninghamia and Sciadopitys have several. All the Abietinew have two ovules only. The relative position of the ovules and their arrangement when reduced in numbers have been alluded to under the head of Sciadopitys. Mention may also here be made of the arrange- ment of the ovules in Callitris and Cupressus. In the young state the ovules are crowded at the base of the fruit-scale, but as the scale grows older it grows in length and the ovules are carried up with it. An examination of the ovules in these plants or of the scars whence they have fallen, seems to show that the arrangement is one of decussating pairs, modified by pressure and compulsory arraugement in one plane: thus two lateral ovules are succeeded by one median one, its fellow being suppressed ; above the median ovule come two more lateral ones, and so on. In the Sciadopitys just alluded to the arrange- ment was centrifugal or cy mose. As the ovules ripen into seed they sometimes becomes fleshy ; thus the seeds of Ginkgo, Cephalotaxus, and Torreya (figs. 28 and 29) ciosely resemble those of Cycas, there being a fleshy covering enveloping a woody shell, within which is the perisperm Fig. 28.—Seed of Torreya grandis and sections. er (prothallus), covered by a thin membrane adherent beneath to the woody shell, so that on opening the latter the lower part of the Perisperm is bared, while the upper part is clothed with a membranous cap exactly as in Cycas. 324 : DR. M. T. MASTERS ON THE MORPHOLOGY, In other instances the test of the seed expands into wing-like expansions at the top, as in Pinus, Agathis, Ze, or at the sides, as in Thuya, Cupressus, &e.* — - Vi Fig. 29.— Seed-bearing shoot of Cephalotaaus pedunculata, Seeds fleshy, olive-coloured. Dispersion of the Seed.—V arious provisions are met with to serve this end. The fleshy arilof the Yew, the dilated stalk of Podo- carpus, the pulpy seed of Ginkgo and others, the succulent fruit of Juniperus, may all facilitate the dispersion of the seed by birds- .. * See Bertrand, “ Teguments séminaux des Gyinnospermes," Ann. Sc. Nat. 1878), p. 88. ANATOMY, AND LIFE-HISTORY OF THE CONIFERE. 325 The winged seeds also seem to afford ample provision for dispersal. In some cases (Abies) the seeds are liberated by the loosening and falling away of the fruit-scales; in other cases (Picea) the fruit-scales simply separate for a short space, whilst in some species of Pinus the cones remaiu attached to the branches with their thick woody scales firmly closed for an unlimited period, only opening, as it would appear, when the forest-fires cause them to split asunder and liberate the seeds. It is noted that the trees in these forests are all of about the same age and dimensions, a fact accounted for by the circumstances just mentioned. That the seales of the young cone should separate ore from another at the tips to permit of the ingress of pollen is readily intelligible; but the corresponding movements to facilitate the egress of the seeds are not so well known. I have no experience of my own to offer on this subject, but the following extract from a communication of Mr. Albert Prentiss on the Hygro- scopic Movements in the Cone-scales of Abietinee may be read with interest :— * In most of the Abietines, soon after the maturation of the cones, the persistent scales fall backward or outward from the axis to permit the ripened seed to escape. The scales are very sensitive to moisture, and in many species exhibit very rapid movements when wet, as with rain. This is especially well seen in the cones of Tsuga canadensis, in which the widely open scales become completely closed in twelve minutes. This property of the cone-scale is found to be very efficient, first in loosening the winged seeds from the scale which bears them, and secondly in favouring the wide dispersion of the seed as the cones open and close many times before all the seeds are sown, thus securing their transport in different directions by the varying winds” *. In Museums it is a common practice to restore the original form of a cone in which the scales have separated one from the other from drought by soaking it in water. * Prentiss, in * Botanical Gazette’ (1888), pp. 236, 237. 326 ' DR. M. T. MASTERS ON THE MORPHOLOGY, GENERAL REVIEW OF THE NATURE OF THE FEMALE FLOWER. The comparative examination of the female flower under various aspects—morphological, teratological, developmental, and structural—that has now been attempted, though naturally incom- plete, and possibly incorrect in some of its details, permits of a general review being made. In all cases the ovule is the essential part of the flower. This is subtended in all cases by a bract, and is generally borne upon a “ fruit-scale.”” The bract may bea mere scale or it may be leaf-like or woody or succulent, but it has invariably the relative position, disposition, and essential anatomical characters of a leaf. This is admitted by all botanists. The diserepancies and variations of opinion which have given rise to such a voluminous mass of literature concern what is here called the fruit-scale. As this has been throughout referred to incidentally, it is not necessary in this place to occupy space with a detailed account of the views held by each observer, the facts and arguments he adduces in their support, or to increase the length of this communication by citing the corresponding arguments used in their refutation by other observers *. Suffice it to remark that what is here termed the fruit- scale, “ lamina ovulifera,' is almost invariably, perhaps always, present in some guise or other. It is scarcely if at all visible externally in the verticillate Cupressinez, hardly more so in Agathis or Cunninghamia. In other Cupressinese, such as Libo- cedrus, in Araucaria, Sciadopitys, and most Abietinesw it forms the larger and more conspicuous part of the ripe cone. In some genera, as before cited, it is not apparent on the surface; in others it exists as a thin membrane, in others as a spongy substance or a thick woody mass. In some it seems to be quite distinct and separate from the bract; in others it is more or less completely inseparate from it. One constant feature it preseuts wherever development has gone on sufficiently for the dif- terentiation of vascular and bast tissue, and that is that the orien- tation of the vascular or xylem elements and that of the bast or phloem tissue respectively is exactly the reverse of what is met with in true leaves. For the rest it has been considered either * The history of the subject may be found in the several papers of Stras- burger, Eichler, Dickson, Celakovsky, and others before referred to. ANATOMY, AND LIFE-HISTORY OF THE CONIFERÆ. 327 as a leaf or leaves, as a shoot with or without leaves, as an outgrowth from a leaf or from an axis, or as something neither truly foliar nor strictly axial:—thus it has been considered as a leaf (Richard, Lindley); an open carpellary leaf (R. Brown); a ligula, enation, or placental outgrowth from the bract-leaf (Sachs, Eichler) ; a “dédoublement” of the bract (Brongniart) ; a pair of leaves belonging to an otherwise undeveloped shoot and more or less coherent and twisted out of place, a view held with more or less modification of detail by R. Brown, Von Mohl, A. Braun, Caspary, Oersted, Parlatore, Van Tieghem, Stenzel, Willkomm, Engelmann, Celakovsky, and Velanofsky; a disk-like outgrowth from the axis of a leafless shoot axillary to the bract (Stras- burger); a rachis (F. von Mueller); a cladode (Sehleiden, Baillon, Dickson, Arcan geli). It may be observed that these opinions are not really so diverse as they appear to be, for every one of them is sup- ported by some fact or other beyond dispute, yet also by assumptions of more or less doubtful validity and by arguments more or less partial in their application or untenable in the light of more recently acquired knowledge. Embarrassed by such eonflict of opinion, some have tried to evade the difficulty by the assumption that the organ in question has a distinct morphological significance in different genera. The substantial uniformity of anatomical structure throughout the whole series, however, seems to negative this suggestion, and to show that What it is in one member of the order that it is, with more or less modification, in all. While anxious to add to our store of facts I should hesitate to indulge in any hypothetical explanation of them that was materialy at variance with any proposed by more competent botanists. Chiefly because it is scarcely if at all inconsistent with any of them and involves no unproven assumption, I venture to offer the following hypothetical explanation for consideration. The facts detailed in former pages show beyond dispute that the fruit-scale is something superadded to the bract ; that it may arise as an enation either from the base of the bract or apparently from the axis just within or above 1; that its structure is neither that of the leaf proper nor that of an ordinary shoot, but that it does present close resem- blance in structure with a cladode as described by Dickson. 328 DR. M. T. MASTERS ON THE MORPHOLOGY, Reverting for à moment to Casimir de Candolle's * Théorie de la feuille, it may be remembered that this botanist compares the leaf to an axis, the upper half of whose vascular system is abortive or undeveloped, for which reason the xylem is towards the upper or inner surface, the phloem towards thelower. Apply a similar explanation to the fruit-scale, and the position of xylem and phloem becomes intelligible. According to this view the fruit-scale is an enation, either from the bract or from the axis, it is immaterial which, of the nature of a cladode or modified shoot. The lower or outer portion of this branch or cladode is abortive, and consequently the xylem is towards the lower or outer, the phloem towards the upper or inner surface. As the bract-leaf and the fruit-scale are in close apposition in the young state, considerations of space and exigencies of packing in small compass would bring about the reduction or obliteration of the two opposed surfaces, just as in synanthic flowers it usually happens that one or more of the originally contiguous parts is as it were squeezed out of existence. Suppose such synanthy to become hereditary—as the bract and fruit- scale have become—there would then be no present process of obliteration—abortion would have given place to entire suppres- sion, and each organ or member, bract, or fruit-scale, would be defective by reason of the hereditary non-development of some portion of its tissue. Such an explanation of the nature of the fruit-scale appears to me to be consistent with the facts of morphology and anatomy, and not essentially inconsistent with any of the published explanations that are not merely conjectural but based on ascer- tained facts, and moreover it does not necessitate any postulates or unproven assumptions. ANATOMY, AND LIFE-HISTORY OF THE CONIFERS. 329 CONTENTS. Page INTRODUCTION..........ccccccsssecssenescecccccceceeseuseaaececceeseusanceuccececcensusnes 226 THE SEEDLING PA. 230 GERMINATION ......ccccccccscecceeeeee DEE Se Radicle `... 23) Caulicle ...........issseeeseeeeceeeeee eterne nen DEET e Root .........esseeeeeeeesee eene ennneec RM me Cotyledons ............ccccccseccacsecceceecccecscceccecnaeecscsseusesece DE 2i Microscopic anatomy `... wii Plumule MNNNNNNNMNNMNNNMMMMMMMMMMMMMMMMMMMMKNF77IINM 26 FOLIATION SNMMMMMMMMMMMMMMM 2 Parts of the leaves......................-- EEN Se Arrangement of the leaves ............ ccce Së Homotaxy and Heterotaxy EEN Ze Movements of the leaves nemen Sp Microscopic anatomy ................. EEN - Stomata ec EEN 250 Hypoderm ..............eeeseeeeeeeeetee nene nennen hts rennen ^ Mesophyll ................ AEN E ER KAN KEEN Nd ERR eo ses KEREN RENE est en Si Resin-canals ............. esee mene nennt , eeeeee " Endoderm 0.0... ..ccccccccsseeccccseccccececeuvecsececeeecaeasseseeeeeeeseeeees E Periegele ener nnn nn hen nnn tenen Sei Fibro-vascular REENEN eg Bibliography EEN ee Homomorphy and Heteromorphy ............ een 2 Primordial or protomorphic leaves.......... ettet nenne "id Relation to congenital or to external conditions ..................... oe. Leaves on fertile branches ............ nm Sé Concrescence or imperfect Separation. eee oi Arrest or exaltation of Sroeth, En ons Secondary leaves of Pinus | .......... eene Së Anatomical structure .......... eene ennt " : Monophyllous Pines ............... enne "^ Fascicled leaves of CUNNINGHAME EE Si Bups anp BRANCHES 00.0... ..cccccccaccccccnccencccccecceuveceuseeesacseceaesecereeceees "i Arrangement ........ Lesser creen nenne retentis 270 Deperulation 0.00.0... .cccccccccsccesesseseceescesseescesserecesevecesecesceaseneens 2 Order of development of buds EE SS Direction of growth of shoots EN x Shape of the young shoot...... cesses meme emere 2 Spurs RENE MEME oe Cladodes or Phylloclades ENEE a 6 Phyllocladus |... eise mme "i G Sciadopitys. EE SN Anatomy and development ...... DEET 27 i i i DTN í Discussion as to their nature .........«- "m" 330 DR. M. T. MASTERS ON THE MORPHOLOGY, Page RAMIFICATION ENEE ER DEET 280 Fastigiation | ............. DEET 281 Direction of the branches ..............eeeeee enne eene nnne nennen 282 Paucity of buds .................. eee DEE DEE 282 Serpent Firs Ae, REENEN 983 Multiplicity of buds ue 284 Burrs Eu DN 284 Change of direction of shoots EE 284 Adventitious shoots .................... NNNMMMMNMMMMNMMMMMMMMWMWMWMWW—---———-—— 284 Illustrations of different modes of Ramification in Cupressinese, &c.... 286 Tug MALE FLOWERS EEN esee tene 290 Their nature and constitution .............. MNEMMMMMMMMMMMMMMMMM—— 290 Position EEN ennt 291 Number and arrangement REENEN nene ... 293 Perulation ENEE EEN 294 Phyllotaxy TTT 294 Form, size, and colour ......................ee eese EE Di 295 Anthers EEN et tee nennen 295 Number ................ DEEN DEED UU nn 296 Direction ................... e DEE m 2 Dehiscence ...............« DEET ... 296 Connective AE estreno saves 296 Pcllen-grains NEEN 297 Relation to leaves ER 998 Tne FEMALE FLOWERS EE ... 299 Their nature and constitution .......... A NM" 299 'Taxacez and Podocarpes................... eeseuae chess ao teet uuo MT 299 Pherosphera .occcccsceccceccccvccccsccscacsscsccseuvenees MM 299 Saxe- Gothed ER 299 Dacrydium ......... Lese eese eese eene eene nentes teta nent s. 299 KC TT .. 299 Torreya NEEN 300 Cephalotazus .......cccccceccsecee eee eren een BEE 300 Ginkgo oe eccecccccecccccececeusecccuccccucceasceccecs veces ececeesceceee® 300 Phyllocladus ENEE 300 Mierocachrys ....c.cccccececcescccccccucccecncncccaccecsencscssccossesee? 301 Podocarpus Ze EEN 301 Cupressineæ `... DTD 301 Diselma ENEE 302 Callitris ............... see. NEN .. 902 Actinostrobus ............. MEER NEE NEEN 302 Juniperus Lue eee MM .... 902 Thuya ......... eese EEUU DEE 302 Fitzroya En beer MEE 302 Libocedrus ................... eee M MH sees, 002 Thuyopsis. en sooseceesansccsvedess nn OOZ BG s ANATOMY, AND LIFE-HISTORY OF THE CONIFERE. Page Tur FkuALE Frowers : — Cupressus &e. ..... secet nennen then nter te venne 302 Anatomical structure `... 302 Actinostrobus ......... MEER TN 302 Cupressus | ....... esee MEE SS Z2 302 Nature of the Fruit-sealein Cupressinem .................. 303 Taxodiee `... PAREM nn 308 Cryptomeria ..... EEN DEE nn, 004 Athrofazis E VENE 304 Taxodium |... eene tenen hee nennt esses sess sena 304 Sequoia, Ae... «ee 904 Amuenries. EEN 304 WEE 304 Anatomical structure of the bract ...................... e. 306 Agathis RR 307 Anatomical structure... DEEN 307 Araucaria ANNE 307 Sciadopitys....... EEN 307 Anatomical structure, &c., fe, 308 Ábietinem `... eene neetanhheseh ete ene tee enn 308 Pseudotsuga ..... eer T EM 309 Abies e DEEN BEEN ee een een een OUO Picea, &o., Ae, een eene eene enn AEN ee un wun 309 Development of the scales | .................-. NEUE LU Pinus uncinata ........ BREED —-—s: Vets e Saee haa 309 an — TÉCUT UG... eee eee ene et eee neenon me" 310 » pumilio ...... ME See ER EE KREE e... GLO Anatomical structure of bract and scale...... een uneneen GLO Malformations of the Flower ......... ENER ANNE ER EN ENEE suen OLL Arrest of development in Thutecarpus .............-- sectevensssesersss 311 Monacism ...........Leeeeee eee ener E NEM 311 Phyllody of the stamens ............«« ee enee dE eERE KEREN 311 Androgynous Cones ENEE 311 Cupressus Lawsoniana EEN . 912 Picea nigra and P. alba |... eese 313 Pinus rigida and P. T. hunbergii ........ eee 313 Larix europea, |. 313 Proliferous Cones En 314 Cryptomeria japonica |... secet etn . 314 Larix europea....... eese nnne 315-316 Tsuga Brunoniana ENEE 515 Pseudotsuga Douglasii ...... eee sessecseneeeesans 316 Picea excelsa. EEN 318 Eichler's ien... 319 332 DR. M. T. MASTERS ON THE CONIFERS. Prge Malformations of the Flower :— Adventitious buds of Lycopodium and Isoëtes ........++++ 319 Pinus EE 320 Sequoia |... Lesser 320 Sciadopitys........ Leere ehem nennen nme 320 Seedless Cones EE 322 Pinus muricata EE 322 False Cones, Gallen 322 OVULES AND Berns... EEN 322 Arillate and exarillate EE 322 Erect and inverted `... 322 Number EE 323 Arrangement EEN EE 323 Modifications of Coat............ccccccscececcenccceececcsccecesseceseeenaasocseeees 323 Dispersion of the seed ..................... cce 324 Hygroscopic property of the scales, EE 325 GENERAL REVIEW OF THE NATURE OF THE FEMALE FLOWER en 326 What is the fruit-scale, and whence does it spring? .......... sseessoeees 326 Conclusions `... NEE VM" 328 \ REVISION OF THE BRITISH WILLOWS. 333 A REVISION OF THE Baytex WirLows. By F. Bucnanan Warre, M.D., F.L.S. [Read 6th June, 1889.] (Pares IX. XI.) I. INTRODUCTORY. “Hunc locum sibi postulat Salicum familia, que si ulla in Botanicis obscura, bec sane maxime," wrote Linné in the * Flora Lapponica,’ and the unanimity with which all salicologists have echoed the sentiment is eminently suggestive. For the following attempt at a revision of the British Salices I would therefore bespeak a lenient criticism on the part of those interested in the subject. Though the verdict, pronounced some years ago by one of the most eminent of our botanists*, that the “definition and classification of Willows has long been a disgrace to systematic botany,” is still too true, it must not be forgotten that in the early post-Linnean days of botanical science much good work was done by British salicologists. “Full thirty years,” says Sir J. E. Smith, “ bave I laboured at this task ” (of specific definition), “ ten of them under the instructive auspices of my late friend Mr. Crowe, in whose garden every Willow that could be got was cultivated..... The plants were almost daily visited and watched by their possessor, whom no character or Variation escaped ; seedlings innumerable, springing up all over the ground, were never destroyed till their species were deter- mined, and the immutability of each verified by our joint iuspec- tion. This was the more material, to set aside the gratuitous suppositions of the mixture of species, or the production of new or hybrid ones, of which, no more than of any change in estab- lished species, I have never met with an instance." (Eng. Fl. iv. p. 164.) In the work just cited 64 “ species ” of British Willows are described, a number which may be contrasted with the 30 of Babington’s ‘Manual’ (8th edition, 1881), and the 18 of Hooker’s * Student's Flora’ (3rd edition, 1884) ; but since in the latter hybrid forms are not numbered, the comparison with the former is more just. Smith’s Species were not suppressed all at once. The sali- cologists who followed him were rightly so imbued with a con- Sclousness of his great labours, that they were naturally averse * Professor Babington, in Journ. Bot. i. (1863) p. 167. LINN. JOURN.— BOTANY, VOL, XXVII. Za 334 DR. F. BUCHANAN WHITE'S to differ from him without good reason, though some of them * could never satisfy " themselves as to the characters of some of the supposed species. The accompanying Table (Plate IX.) will show, in a perspi- cuous manner, the very different views, as regards the number of British Willows which have been held at various times. The shaded columns show the quantity of numbered “species” at each period. The darkest portion indicates the number of true species included in each estimate; the medium the number of what are now (in this paper) recognized as hybrid forms ; and the lightest the number of supposed species, but which are now considered to be forms only, or, if species, have no claim to be regarded as natives of Britain. The periods are as follow :— 1762. Hudson, ‘Flora Anglica! 18 species (10 true, 1 hybrid). Hudson's S. reticulata is a form of S. herbacea. 1804. Smith, ‘ Flora Britannica,’ vol. iii. 45 species (16 true, 8 hybrids). 1828. Smith, ‘English Flora,’ 2nd ed., iv. 64 species (17 true, 6 hybrids). 1830. W. Withering, Tun. Arrangement, 7th ed. 52 species (17 true, 6 hybrids). . 1835. Lindley, ‘Synopsis, 2nd ed., and 1841, 8rd ed. 30 ~. species (17 true, 5 hybrids). 1888. W. J. Hooker, ‘British Flora, 4th ed. 70 species (17 true, 18 hybrids). 1860. Hooker and Arnott,‘ British Flora,’ 8th ed. 38 species (17 true, 6 hybrids). 1878. Boswell, ‘English Botany,’ 3rd ed. 34 species (17 true, 10 hybrids). 7 ' 1881. Babington, ‘Manual,’ 8th ed. 30 species (17 true, 7 hybrids). , 1884. J. D. Hooker, ‘Student’s Flora," 3rd ed. 18 species (16 true, 1 hybnd).—In this work supposed hybrids, though mentioned, are not numbered ; and S. cinerea 18 treated as a subspecies. 1886. ‘London Catalogue, 8th ed. 31 species (17 true, 7 hybrids). NEM 1890. As set forth in this paper, 68 forms (17 species, 4l hybrids). mM E cht Wa EE E REVISION OF THE BRITISH WILLOWS. 335 From this statement it will be seen how great the range of opinion has been during the present century, the estimated number of “ species " varying from 70 in 1838 to 18 in 1884! But though since 1838 so many names have disappeared as the names of species, many of them are not only still retained in our lists as those of varieties, but have been added to; so that in the last edition of the * London Catalogue' there are no less than 96 named willow-forms included under 31 species. The object, therefore, of this Revision is in the first place to con- sider how many of these names deserve retention; and in the second to point out the occurrence in Britain of some hitherto unrecorded hybrids. Ín endeavouring to carry out this intention it has happened that I have had occasionally to criticise the work—either the descriptions cr the determinations of specimens—of some of the great salicologists ; and I wish to state, though I daresay it is unnecessary, that such has been done in no carping spirit. For the sake of brevity, I have not cited at length under each species the works chiefly consulted, but have referred to them simply by the names of the authors. The abbreviations thus used are as follow :— Andersson. (N.J Andersson, * Monographia Salicum,’ 1867.) —Of this great work, which, with Wimmer’s * Salices, is of the utmost importance to the student of Willows, the first part only was published. The remaining species, as well as some rather later views of the author, are given in De Candolle's * Prodromus,’ Pt. xvi. section 2, 1868; and for all species after and inclusive of the Viminales the citation “Andersson” refers to the ‘ Prodromus.’ Other works of the same author which have been consulted are the ‘Salices Lapponim, the genus Salix in Blytt's ‘Norges Flora,’ pt. ii. 1874; and the notes made on Leefe’s ‘ Salictum Britannicum, and communicated by H. C. Watson to the ‘ Botanical Gazette, May 1851. u Babington. (Professor C. C. Babington, ‘ Manual of British Botany,’ 8th ed., 1881.) Boswell-Syme. (J. Boswell-Syme, ‘ English Botany,’ 3rd ed., 1873.) . Forbes. (James Forbes, ‘Salictum Woburnense,’ 1829.)* * For a large number of living specimens (from plants cultivated at Kew and probably originally derived from the Woburn collection), illustrating the *pecies figured by Forbes, I am greatly indebted to Mr. George Nicholson, 242 336 DR. F. BUCHANAN WHITE'S J. D. Hooker. (Sir J. D. Hooker,‘ Student's Flora,’ 3rd ed., 1884.) W. J. Hooker. (Sir W. J. Hooker, ‘ British Flora, 4th ed., 1838. Borrer's opinion is frequently quoted in this work.) Koch. (W.D. J. Koch, * Synopsis Flore Germanicæ, 2nd ed., 1844.) Smith. (Sir J. E. Smith,‘ English Flora, 2nd ed., 1828.) Walker Arnott. (Hooker and Arnott’s ‘British Flora, 8th ed., 1860, for which Dr. Walker Arnott was alone responsible.) Wimmer. (F. Wimmer, * Salices Europææ, 1866.) Any other works referred to (and many have been consulted) are cited at greater length. In addition to consulting the descriptions, I have, when pos- sible, compared our specimens with the examples published by several salicologists, and more especially by Wimmer. For an opportunity of doing so with great convenience, I am much indebted to the kindness of Mr. C. Bailey, who ient me his extensive collection of Willows, which includes, in addition to Wimmer's ‘Collectio Salicum Europzarum,’ many examples pub- lished by A. and J. Kerner, Reichenbach, &c. Other specimens examined include Linné’s own herbarium, belonging to the Linnean Society *; Smith’s, also in the posses- sion of that Society; some specimens of Smith's attached, with notes, to the original drawings for ‘English Botany,’ pre- served at the British Museum; the British Museum and Kew Herbariums ; the British Willows of Edinburgh University Herbarium (kindly lent me by Professor Bailey Balfour) ; and several private herbariums, some of which contain, in addition to the specimens published by the Rev. J. E. Leefe, other examples received from that botanist. Mr. Leefe's specimens are valuable, not only as showing his own and Ward's opinion on many British Willows, but as illustrating the species of Smith and the views of Borrer. Besides the specimens just mentioned, I have examined in à living condition several thousand examples, either collected by myself, or sent to me by friends and correspondents, to whose kind services I am much indebted, and whose names are men- tioned under the species of which they have provided me with much-needed specimens, either living or dried, * For information regarding the Jabels of this collection I am greatly in: debted to Mr. Daydon Jackson. REVISION OF THE BRITISH WILLOWS. 337 Il. CLASSIFICATION. Various arrangements of the Willows have been proposed; but perhaps the best is that of Andersson, though even that is not altogether satisfactory. Andersson arran ges the species in three tribes :—A. Pleiandre, B. Diandre, C. Synandre. The essential characteristics of the Pleiandre lie in the pale unicolorous scales, which fall off before the fruit ripens; in the stamens being most usually not less than three in number; and in the nectary being commonly double. The Diandre are distinguished by having the scales, which are more or less darker-coloured in the upper part, persistent with the fruit ; the stamens two in number and with free filaments; and the nectary very rarely double. The Synandre may be recognized by the filaments of the two stamens being more or less connate. Now whilst these points are, in the main, characteristic of each tribe, they are not such decided distinctions as might be desired. In the Pleiandre, for example, the stamens, though described as free, are not so in every species; since in some, as in S. fragilis, they are, or appear to be, connate at the very base When viewed from the back—from the front the adnate anterior nectary conceals the union; the nectary, though often double, is in one sex of some species single only ; and the stamens in some species are almost constantly two in number. Moreover, the scales in some species of both the other tribes are unicolorous. Andersson subdivides his three tribes into sections, and these again into groups. In the Pleiandre there are two sections— the Tropice, with four groups, and the Zemperate with three, namely, 5, Amygdaline or Triandre, 6. Lucide or Pentandre, and 7, Fragiles or Albe. The Diandre are divided into the Microstyle (which includes 8. Longifolie, 9. Cinerascentes or Capree, 10. Rosee, and 11. Argentee or Repentes); the Podo- style (12. Virescentes or Phylicifolie, and 18. Rigide); and the Macrostyle (14. Pruinose, 15. Micantes or Viminales, 16. Nivea, and 17, Nitidule). The Synandre have two groups only— 18. Incane, and 19. Purpuree. Adopting Andersson’s arrangement, but using in some cases 18 second name for the group as being more instructive, the British Willows may be classified as toilows :— 838 DR. F. BUCHANAN WHITE’S A. PLEIANDRZE. 1. TRIANDRZ. 1. Salix triandra, L. x Salix decipiens, Hoffm. x Salix subdola, B. White. x Salix undulata, Ehrh. 2, PENTANDRE. 2. Salix pentandra, L. x Salix cuspidata, Schultz. x Salix hexandra, Ehrh. 8. FRAGILES. 8. Salix fragilis, L. b. britannica, B. White. 4. Kalix alba, L. b. vitellina, L. x Salix viridis, Fr. B. DIANDRÆ. 4. CaPRER. 5. Salix cinerea, L. 6. Salix aurita, L. x Salix lutescens, A. Kern. 7. Salix Caprea, L. x Salix Reichardti, A. Kern. X Salix capreola, J. Kern. 5. REPENTES. 8. Salix repens, L. x Salix ambigua, Ehrh. x Salix cinerea-repens, Wimm. x Salix Caprea-repens, Lasch. x Salix nigricans-repens, Heidenr. 6. PHYLICIFOLIZ. 9. Salix phylicifolia, Ta. a. S. phylicifolia, L., auct. b. S. nigricans, Sm. c. S. phylicifolia-nigricans, Wimm. REVISION OF THE BRITISH WILLOWS. 389 x Salix laurina, Sm. x Salix Wardiana (Leefe, MS.), B. White. x Salix ludificans, B. White. X Salix tephrocarpa, Wimm. x Salix latifolia, Forbes. x Salix strepida (Schleich.), Forbes. x Salix coriacea (Schleich.), Forbes. 10. Salix Arbuscula, L. x Salix Dicksoniana, Sm. 7. VIMINALES. 11. Salix viminalis, L. x Salix Smithiana, Willd. a. stipularis (Sm.). b. sericans (Tausch). e. velutina (Schrad.). d. ferruginea (G. And.).. e. acuminata (Sm.). 8. NIVES. 12. Salix lanata, L. b. Sadleri (Syme). X Salix superata, B. White. X Salix Stephania, B. White. 13. Salix Lapponum, L. b. helvetica (Vill.). x Salix aurita- Lapponum, Wimm. X Salix cinerea-limosa, Læstad. x Salix spuria (Schleich.), Willd. 9. NiTIDULE. 14, Salix Myrsinites, L. x Salix Wahlenbergii, And. X Salix saxetana, B. White. X Salix serta, B. White. ` 15. Salix herbacea, L. x Salix Grahami (Borr.), Baker. X Salix Moorei, “ Watson, L. C." x Salix simulatrix, B. White. X Salix sobrina, B. White. X Saliz margarita, B. White 840 DR. F. BUCHANAN WHITE'S 16. Salix reticulata, L. x Salix semireticulata, B. White. x Salix sibyllina, B. White. C. SYNANDRA. 10. PuRPUREX. 17. Salix purpurea, L. X Salix rubra, Huds. X Salix sordida, Kern. X Salix dichroa, Doll. x Salix Doniana, Sm. It will be noticed that in this list one willow, generally included in British catalogues, has been omitted. This is S. daphnoides, Vill., which, though not unfrequently planted, can in no way be claimed as an indigenous plant. It will also be noticed that in a very few cases only are varie- ties distinguished by name. “ Varieties,” “forms,” and “ modi- fications” of almost every species and hybrid have at one time or other been described; but since the forms so separated have, iu the vast majority of cases, no constancy, but pass by imper- ceptible gradations the one into the other, their retention is & hindrance rather than a help. In the few instances where I have distinguished varieties, these have not all quite an equal rank; but their value will be indicated as each of them is specially discussed. I may also state that I shall only occasionally allude to gynan- drous forms, though in many cases names have been bestowed on them. A number of species have not very unfrequently a mixture of male and female flowers in the same catkin, whilst others have their floral organs quite monstrous. Such forms, though of con- siderable interest, ought not to be distinguished by name. III. HYBRIDIZATION. In addition to the really great variability of the true species and consequent difficulty in attaining a satisfactory knowledge of them, the fact that Willows hybridize with the greatest facility adds immeasurably to the intricacies of the study. The earlier salicologists were mostly unwilling to recognize the pos- REVISION OF THE BRITISH WILLOWS. 341 sibility of hybridization, though such had been suggested ; for Smith writes of “the gratuitous suppositions of the mixture of species, or the production of new or hybrid ones.” At a somewhat later period, however, the probability of the phenomenon was admitted ; but it was left to Max Wichura * to prove, by experiment, the truth of what had before been only— though on good grounds—suspected. Wichura found that not only did many binary hybrids occur naturally, but that ternary hybrids were also spontaneously produced. Binary hybrids are those produced by two species; ternary are those into whose composition three species have entered. In addition to this, he showed that by cross-ferti:izing these hybrids, plants could be produced whose pedigree included no less than six species. (The accompanying diagram (Plate X.) shows graphically the pedigree of the compound hybrids.) Six species appeared to be the limit, as, from the imperfection of the pollen or of the seeds, the combination of species could not be carried further. Theoretically, therefore, every species can form spontaneous hybrids with every other species; but practically hybrids are restricted by several causes. For hybrids to occur, not only must there be a certain degree of proximity of situation in the parents, but an identity in the period of flowering. Close Proximity, though favourable for hybridization, is not absolutely necessary, since fertilization is accomplished by insect agency, and the attraction to insects of willow-flowers is very great. Identity of the usual time of flowering is an almost imperative necessity for the production of hybrids; but these also occur— though more rarely—between species which do not ordinarily flower at the same time, and must have been produced by some abnormality in the period of one of the parents. It must also be remembered that species, whose distribution is both lowland and alpine, have their time of flowering retarded in alpine situa- tions, and are hence able to hybridize with the true mountain Species, A hybrid in its best condition is exactly intermediate in cha- racter between its two parents; but more frequently it shows a Sreater relationship with one rather than the other; and in those cases where it occurs in any abundance, a series of speci- * ‘Die Bastardbefructung im Pflanzenreich erläutert an den Bastarden der Weiden; 1865. 342 DR. F. BUCHANAN WHITE'S mens can usually be obtained exhibiting a more or less perfect gradation from one parent to the other. Some of these are pro- bably due to the fact that they are really crosses of the hybrid with one of the parent species ; but this is not necessarily the ease, since the influence of one parent may be stronger than that of the other; and differences may also result from an alteration in the sex of the parents. Thus, if A and B represent two species, then Ad x B9 may, it is supposed, produce a some- what different-looking hybrid than A? x Bd does. To recognize a hybrid, the student must, in many instances, have an intimate acquaintance with the characteristics of the true species, more especially in the case of closely allied ones. This is very essential, since the books usually describe the more distinet forms only, and frequently seem to ignore the insensible gradations which connect the hybrid with its parents: not that they really ignore them, but from the difficulty of expressing in words characters that the trained eye can more or less easily perceive. The rank to be ascribed to hybrids and the system of nomen- clature to be adopted are points on which it is desirable that more unanimity and uniformity should obtain amongst botanists. Taking Nyman's * Conspectus’ as a sample of a not uncommon method, it will be found that some hybrids (e. g. S. rubra and S. Doniana) are given full rank and numbered as species; whilst others, of equal importance, have no such position. Some other authors merely indicate the occurrence of hybrids and do not describe them; whilst among those who give hybrids a rank nearly or quite equal to species, some, as Andersson, place them in the groups to which they are most entitled to belong, but others, as Wimmer, keep all the hybrids together. Then as regards the nomenclature, Andersson and many others adopt distinct names, which do not in any way indicate the real or supposed parentage; but Wimmer, on the other hand, re- pudiates, as a rule, all such names, and uses for the hybrids a combination of the parental designations. In many respects there is much advantage in Wimmer’s method; but even he bas not been able to employ it uniformly, and was obliged to use other names in certain cases where the parentage is doubtful ; and moreover in some instances the compound name that he has used is erroneous and misleading. "Wimmer's plan also entails à breach of thelaw of priority. "Under these circumstances, whilst REVISION OF THE BRITISH WILLOWS. 343 there is a decided advantage in employing a compound name— since it conveys distinct information—such can be used in those cases only where no earlier name exists, and where there is no doubt about the parentage. Regarding the Rank and Position of Hybrids.—In the first place, they can scarcely be treated as mere varieties—much less ignored—if Willows are to be satisfactorily studied; on the other hand, neither can they be considered as equal to species. The title they should bear is that of “ hybrid,” and their ordinal posi- tion should be near the species to which they are allied. As to whether they should or should not be numbered, that is a matter of convenience ; but their names should always have the prefix x, as indicating the hybrid origin. Since, theoretically, every species of Salix cau hybridize with every other species, the 18 British species (giving S. nigricans specific rank for this occasion) should produce about 144 binary hybrids. As a matter of fact, about 61 real or supposed hybrids of these 18 species are known ; but somewhere about 20 of these have not yet been detected in Britain. The species with which each of the 18 hybridizes are as follow :— Triandra, with fragilis, alba, cinerea, aurita, Caprea, ani viminalis. Pentandra, with fragilis, alba, nigricans, and Arbuscula . Fragilis, with triandra, pentandra, and alba. Alba, with triandra, pentandra, and fragilis. Cinerea, with triandra, aurita, Caprea, phy licifolia, nigricans » repens, viminalis, Lapponum, and purpurea. Aurita, with triandra, cinerea, Caprea, phylicifolia, nigri- cans, repens, viminalis, Lapponum, Myrsinites, herbacea, and pur- purea. Caprea, with triandra, cinerea, aurita, phylicifolia, nigricans, repens, viminalis, Lapponum, and purpurea. Eepens, With cinerea, aurita, Caprea, phylicifolia, nigricans, viminalis, Lapponum, and purpurea. Phylicifolia, with cinerea, aurita, Caprea, nigricans, Arbuscula, repens, viminalis, Myrsinites, and herbacea. uu Nigricans, with pentandra, cinerea, aurita, Caprea, phylici- Jolia, repens, Lapponum, Myrsinites, herbacea, r eticulata, and Purpurea. Arbuscula, with pentandra, phylicifolia, Lapponum, Alyrsinites, herbacea, reticulata, and purpurea. 844 DR. F. BUCHANAN WHITE'S Viminalis, with triandra, cinerea, aurita, Caprea, repens, phyli- cifolia, and purpurea. Lanata, with herbacea and reticulata. Lapponum, with cinerea, aurita, Caprea, nigricans, Arbuscula, repens, Myrsinites, and herbacea. Myrsinites, with aurita, phylicifolia, nigricans, Arbuscula, Lap- ponum, and herbacea. Herbacea, with aurita, phylicifolia, nigricans, Arbuscula, lanata, Lapponum, Myrsinites, and reticulata. Reticulata, with nigricans, Arbuscula, lanata, and herbacea. Purpurea, with cinerea, aurita, Caprea, nigricans, Arbuscula, repens, and viminalis. Whilst the foregoing list will serve to show the student what hybrids may be expected to occur, the accompanying diagram (Plate XI.) demonstrates the relation of both sections and species as regards hybridization. The larger circles indicate the sections, the smaller circles the species, and the lines connecting the latter show that these hybridize. As regards Britain, both the Pleiandre and the Synandre are lowland (i. e. not ascending above 1000 feet) in their altitudinal distribution ; but the Diandre include both lowland and alpine species; and the groups thus formed are indicated by dividing the circle by dotted lines and numbering the segments. I. includes S. viminalis only, which is strictly lowland. II. formsa group of which the species, though most common in the lowlands, ascend into the region of III., the strictly alpine species, rarely, if ever, descending to 1000 feet. IV., including S. Lapponum and S. Arbuscula, ought not perhaps to be separated from IIL, since these are very rarely otherwise than alpine in their distribution. Though nine or more ternary hybrids occur spontaneously in Europe, none have been detected with absolute certainty in Britain, perhaps from the great difficulty attending their re- cognition. IV. Corrrerriwae. Since Willows produce their flowers either before the leaves or when the leaves are only young, and since mature leaves are necessary for the proper determination of the species, leaf- specimens and flower-specimens cannot be obtained at the same time; hence, unless great care be taken, there is a real danger that the flowers of one bush and the leaves of another may be taken to represent one specimen. ` REVISION OF THE BRITISH WILLOWS. 345 In collecting, therefore, it is essential to guard against any intermixture of specimens. To do so, not only must the trees be marked, but the specimens taken from them be ticketed in sucha manner as to prevent the possibility of mistakes. The method which experience has shown me to be a good one is as follows :— Provide a number of slips of paper 5 or 6 inches long by 4 inches wide, each with a long slit in it. On selecting a bush from which to take flower- or leaf-specimens, cut on the bark a number in Roman numerals *, put the same number on one of the slips of paper, and add a description of the situation of the bush; then having taken the Specimens required, pass their ends through the slit and transfer to the vasculum. On returning home, enter in the “ locality note book ” the number and other particulars of the bush, and prefix in Arabic numerals the note-book or per- manent number. Write the latter on small bits of paper and fix one to every specimen before it is put in the press to dry. In this way all risk of confusion of specimens will be avoided. Whilst the note-book numbers must run continuously, the bush-numbers can, for the sake of convenience, be repeated when the localities from which the specimens are obtained are suffi- ciently distinct. Before pressing the specimens it is desirable to enter in the note-book such particulars of the plant as can be better seen in the living than in the dried condition. The importance of these will vary according to the species or group. Jn the Synandre, for example, the extent of connation of the filaments and colour of the anthers (at different ages) are points to be noted in the living plant. The colours of the different parts, structure of the style, stigmas, and nectary, venation of the leaves (whether raised or impressed), surfaces (smooth or wrinkled), the margin (flat or ineurved) are included in the characters which should be recorded in the note-book. In drying, pressure sufficient, both in weight and continuance, to keep the leaves flat, without erushing the catkins too much, should be given. Some of the leaves, at various parts of the specimen, should be arranged so as to show the under surface. In selecting leaf-specimens side branches as well as terminal * Herr Hauptmann Schambach, of Northeim, Hanover, has kindly suggested another method of preserving the identity of the bushes. He uses narrow strips of lead (stamped with a number) which can be twisted round a branch. 346 DR. F. BUCHANAN WHITE'S shoots should be taken. The latter alone are not sufficient, as they often have leaves somewhat different from the normal con- dition; on the other hand, they are useful for illustrating the stipules. Leaf-specimens should not be taken till the leaves are mature (since the young leaves are deceptive), and, if possible, not earlier than the middle of August. Male catkins should exhibit both unopened and opened anthers. Female catkins should be neither too young nor too old, but be just about the age of fertilization. Specimens in young fruit are also useful. In studying Willows “a practised eye is,” as Mr. Leefe says, “more to be relied on than the characters found in books; since, as Fries remarks, “Characteres non specierum sunt criteria, sed ad species dignoscendas adminicula. Ex his modo species agnoscuntur, ex vegetationis cognoscuntur.... Hine Linnzus in speciebus discernendis non characteres sed oculorum judiciique aciem laudat." Whilst all parts of the plant are variable, some characters, on which a good deal of reliance has been placed, are so inconstant that they may, in many cases at least, be almost or quite ignored, though in other instances they are really of importance. Fami- liarity with the species can alone teach the student what are the points on which he can depend. In many species the presence or absence of stipules, and the shape of these organs, are of no great value for the discrimination of the plant; in others it is the very reverse. The presence or absence or the amount of pubes- cence is a character of similar value, as is the presence or absence of glaucosity. V. DISTRIBUTION. Though it is probable that the records of distribution in Britain of the true species are, on the whole, correct, the range of the hybrids has yet to be worked out; and for the sake of accuracy it is perhaps expedient that a new census of the distri- bution of all the British Willows should be taken. Perthshire is as rich as—probably richer than—any other county. All the seventeen true species and thirty-one of the hybrids occur in it; and since the neighbouring county of Forfar is probably nearly as productive, and has several hybrids which have not yet been detected in Perthshire, Central Scotland appears to be the metropolis of Willows in Britain. REVISION OF TITE BRITISH WILLOWS. 347 VI. Reviston or rue SPxcr«us. A. PLEIANDRA. Group 1. TRIANDR Æ. 1. SALIX TRIANDRA, L. Smith and his school of salicologists considered that the willow which is now known as S. triandra, included three, if not four, distinct species, viz. S. triandra, L., S. amygdalina, L., S. Hoff- manniana, Xm., and perhaps S. contorta, Crowe,— distinguished by the shape, size, and colour of the leaves, and nature of the shoots. Continental botanists, on the other hand, while not usually recognizing all these, separated from 8. triandra several other forms. At the present day none of these forms are recog- nized as species, though most of them are retained as varieties, at least by the British school. Thus, in the last edition of the * London Catalogue’ four varieties of S. triandra are given, namely, a. amygdalina (L.), b. Hoffmanniana (Sm.., c. Trevirani (Spreng), and d. contorta (Crowe). The majority of continental salicologists, as Koch, Grenier, and Andersson, make the more important varietal characters lie in the colour of the leaves, while Wimmer considers the shape to be of more value. Andersson's leaf-varieties are a. discolor, Koch (underside of leaves intensely glaucous), and f). concolor, Koch ( paler but not glaucous). Of each of these there are the forms 1. latifolia, 2. angustifolia, and 3. microphylla. He also gives two catkin varieties, y. tenuijulis and ò. crassijulis. Wimmer's varieties are a. vulgaris, 3. angustifolia, aud y. Vil- larsiana, each with modifications according as the underside of the leaves is green or glaucous. On the whole, Andersson’s division into discolor and concolor seems to be the arrangement most worthy of retention ; since an examination of any large series will show that the variations dependent upon shape, whether of the leaves or of the catkins, are all connected by intermediates and glide the one into the other. But the presence or absence of glaucosity is not a suffi- ciently stable character, either, upon which to found varieties ; for, as Wimmer remarks, not only may glaucous and green leaves be found on the same plant, but even the same leaf may be partly green and partly glaucous below. It seems better, therefore, to place little importance on any of the so-called varieties—either British or continental—of S. triandra. 348 DR. F. BUCHANAN WHITE’S In Britain, judging from the specimens I have seen, Salix tri- andra exhibits fewer extreme forms of variation (though variable enough both in leaves and catkins) than it does in continental Europe. The ‘ London Catalogue’ var. a. amygdalina seems to be equivalent to Andersson’s a. discolor, which he says is S. amyg- dalina, auctorum (S. amygdalina, L., if not altogether dubious, is only a synonym of S. triandra); b. Hoffmanniana is D. concolor, 1. latifolia of Andersson; and c. T'reviran? is a hybrid of S. tri- andra and S. viminalis, and will be considered hereafter. Speci- mens named by Leefe S. contorta, Crowe, do not altogether agree with Smith's description of that plant, which he says has leaves half the size of those of S. triandra. Leefe’s contorta has long leaves, which in several ways suggest a cross of triandra with fragilis ; but the catkins ( 9 ) are in all essential particulars those. of triandra, of which species it seems to be only a leaf-form. Andersson is of opinion that his discolor is in Western Europe a more truly wild form than concolor, which is the more fre- quently eultivated one. In Britain concolor seems to be the commoner, and is more usually broad-leafed than narrow-leafed. In Linné's herbarium there are not any specimens named S. triandra ; but a plant labelled by Linné “ hastata ” (and by Smith * triandra ? ") appears to be S. triandra, with leaves glau- cous below; and another example, labelled by Linné “ Salix pentandra " (to which Smith has put a “ ? "), and with, in another band, a stuck-down label with “ Salix pentandra, Flor. Lap. 370. Foliis subtus cinereis. No. 8,” seems also most probably S. tri- andra, with leaves narrowed at each end and glaucous below. X SALIX pECIPIENS, Hoffm. (S. triandra x X. fragilis.) In his ‘ Historia Salicum” (vol. ii. fasc. i. p. 9, t. 31, 1791), Hoffmann describes and figures a willow, which he named S. decipiens, from the resemblance of the leaves to those of S. bigemmis (=S. daphnoides), and which, he says, is one of several species that go by the name of S. fragilis. Smith, amplifying the description, adopted Hoffmann’s species, and was followed for a considerable time by British botanists. The more recent British authors have, however, placed S. decipiens as è variety of S. fragilis, and have omitted more or less to notice important points of its characteristics, The continental salicologists, on the other hand, very 800! began to consider S. decipiens as either a synonym (e. g. Willdenow, as also Lindley amongst British botanists) or as a mere variety REVISION OF THE BRITISH WILLOWS. 349 (e. g. Koch) of Salix fragilis. Wimmer, while citing the name as a synonym, adds, in his notes, that the plant he has described is the same as Hoffmann has figured, and that, though no adult leaves are depicted, yet it is evidently a broad-leaved form, which he considers to be the type of the species (i. e. of S. fragilis). Andersson dismisses S. decipiens as only a slight modification of S. fragilis, analogous to the var. vitellina of S. alba, and pro- duced, like that, by the annual lopping of the tree. The short description he gives is similar to that given by Koch. A careful study of the figures and descriptions will, I think, suggest that the plant attributed by Andersson &c. to S. deci- piens is not Hoffmann's species, but a modification only of S. fragilis, produced, as Andersson says, by annual cutting-over, and distinguished by the bark of the twigs being testaceous in colour, and the lower leaves more or less obtuse. No allusion is made, it will be noticed, to the remarkable “ polish " of the twigs, nor to the inflorescence. In Britain, however, the plant described and figured by Smith as S. decipiens (and which is beyond doubt Hoffmann's species) has been continuously known under that name, though the more recent writers have, in reducing it to the rank of a variety, omitted to notice some of its essential peculiarities, and have in fact apparently not observed them. As a matter of fact, Hoffmann’s decipiens seems to be little known and scarcely understood by continental salicologists. Andersson saw in H. C. Watson's herbarium * specimens pub- lished by Leefe t, and made no remark upon them, except that there were at Upsala two trees, planted by Linné, altogether like Leefe’s plant, and that the form wasvery rare in Sweden. On this note Leefe and Ward make the following comment (Journ. of Bot. Viii. p. 305): * certainly decipiens, E. B." Possibly the specimen (which, though in bad condition, seems to be quite the same as the British decipiens) in the British-Museum Herbarium, labelled " Saliz Sragilis, var. decipiens, Koch; Upsala; E. Fries (Herb. Norm.),” may be from one of these trees. Another specimen * Thave a Willow from Surrey, labelled by Watson “ Salix undulata, fide Andersson,” which, though it is a leaf-specimen only, I have no doubt is S. decipiens, t No. 50 of original fasciculus. Leefe says, “ buds black in spring”; but his Specimens have pale buds. He also says, “ In specimens received from Professor Koch the scales of the 9 are round and very hairy; the leaves of the d are broader and less glaucous and reticulated beneath than in my specimens ; those of the © agree exactly.” So apparently Koch knew the plant. LINN. JQU&N.— BOTANY, VOL. XXVII. 2B 850 DR. F. BUCHANAN WHITE's (d), in Kew Herbarium, collected by Fries at Upsala, and labelled Salix decipiens, Koch, seems to be also the same as our decipiens, but with leaves more in the direction of fragilis. As corroborating the belief that our decipiens is not known, under that name, to Continental botanists, it may be mentioned that specimens quite the same as it were collected at Königsberg, and published as S. fragilis var. porcellanea by Baenitz. These Königsberg examples again bear a resemblance so sufficiently close as to suggest identity of species with a willow gathered near Hanover by Beckmann, and named S. fragilis X triandra, f. androgyna*. From the study of a large number of specimens, living and dried, I have come to the conclusion that S. decipiens is a form distinct from S. fragilis, and suspect that it is a hybrid between S. triandra and S. fragilis. Like all hybrid forms (and indeed all willows) its characters are more or less unstable; but usually it is so well marked that it is not difficult to recognize, though some leaf-specimens may be confounded with S. fragilis and others with S. triandra. In fact examples of what seems certainly to be decipiens were sometimes named triandra by the older botanists, as Forster, Borrer, &c. ` and I have seen bushes and specimens which, so far as the leaves go, could not at first sight be readily distinguished ‘from that species. Comparing typical examples with S. fragilis (since it is with that species that it has been confounded), it will be found that, whereas fragilis attains the stature of a big tree, decipiens does not grow to more than a large. bush or small tree—I refer, of course, to plants which have never ‘been cut over. The bark of the trunk and older branches seenis ‘rougher and more broken than in fragilis, and, though not deciduous, as in triandra, appears inclined to split up. The branches are more upright in their direction than in fragilis (i. e. they forma more acute angle with the stem), and while somewhat brittle at their points of insertion, are less so than in fragilis. "The year-old twigs are, a8 described by Smith, &c., highly polished, shining “like porcelain, as if var- nished ” (had Baenitz this description in view when he named bis specimens var. porcellanea ?), and most usually of a yellowish- white or clay-colour. The polish of the twigs has always been regarded as an important feature; and Smith tells Sowerby, in 4 note on the original sketch of his plate, to show the varnishing well, as it is the chief character.’ The shoots of the year are * Some other Continental examples are simply named S. fragilis. REVISION OF THE BRITISH WILLOWS. 351 often, but by no means invariably, of a fine crimson colour (frequently on the exposed side only), but not much, if at all, polished, and, sometimes at least, longitudinally furrowed. The leaves are constantly smaller than those of Salix Sragilis, and while in the same specimen a good deal of variety of form may be found, they are, as arule, more oblong, more parallel-sided (¢. e. of nearly equal width for a greater part of their length), usually less narrowed, and often indeed rounded at the base, and more abruptly and less longly acuminate at the apex. On the whole, the leaves are broader in proportion to their length than in fragilis. In colour and texture there is a distinct difference, but one more readily seen than described. The upper surface is less shining and of a duller green, and the underside is pale dull green, closely reticulated all over, from the veins, even the smallest, being dark green. The secondary veins (those springing from he midrib) in the larger and broader leaves seem to form with the midrib an angle more acute than in triandra, and less acute than in fragilis. The serration is also more irregular than in fragilis. In the ordinary form the leaves are always glabrous (except perhaps at the very first), nor are they glaucous below; but in the forms nearer fragilis there are exceptions to this. l Comparing the ¢ catkins with those of fragilis, they will be found to be quite unlike those of the common British form, inas- much as they are denser-flowered and have the filaments much longer than the scales. They are much more like those of the European form, but are constantly smaller, and not so thick in Proportion to their length. The catkins seem to be always fewer in number on a twig than in fragilis, and much fewer than in triandra ; and, so far as my observations go, the flowers are most usually diandrous, but sometimes, though more rarely, triandrous. Iam not sure that any important characters lie in the peduneular leaves, Which are variable and usually entire. The 9 plant seems to be scarcer than the d; aud, indeed, Boswell Syme says that the d only is now known in Britain, which is not, however, the case. Hoffmann describes the ovary as attenuate from an ovate base, stalked, with the style scarcely distinct ; and the capsule as oblong acuminate from an ovate base. Smith says the ovary is lanceolate on ashortish stalk, and tapering into a stout style one third its own length, and that the stigmas are half as long as the style. Smith does not ngure the Ovary; and from Hoffmann’s figure there might be said to be either no style, or else one one-third the length of the ovary so 2522 852 DR. F. BUCHANAN WHITE'S much is it a matter of opinion where the ovary ends and the style begins. In the specimens which I have seen I cannot say that the style is so long as one-third of the ovary, while it is not very markedly different in length from the stigmas, being sometimes & little longer and sometimes a little shorter than these. In typical specimens the ovary, which is smaller than in Salix fragilis, with shorter style and stigmas, is ovate, tapering, but rather blunt (i. e. as it ripens, contracted into the style), with a rather stout style (often bifid) and rather broad spreading or recurved cloven stigmas. The pedicel is about twice the length of the inner nectary, the outer nectary being very small and obscure. In both sexes the scales are variable, but are perhaps, on the whole, less hairy than in fragilis, and more hairy than in triandra. Though for the above comparative description typical (4 e. intermediate) specimens have been selected, it must be kept in mind that many examples diverge either in the direction of fragilis or of triandra, the variation being chiefly noticeable in the leaves. Thus the leaves may be distinctly glaucous below, more distinctly and persistently silky when young, and more Sragilis-like in shape; the bark, especially of the flowering twigs, may be darker in colour than usual ; and the pedicel of the ovary may be three or four times as long as the nectary. Compared with S. triandra, the more polished bark of the year- old twigs, the narrower more acute stipules (when these are present), the diandrous flowers, and the usually shorter pedicel and more distinct style will generally distinguish the plant; but, from the leaves alone, I would sometimes hesitate before affirming positively whether certain specimens belonged to decipiens oT to triandra. I have described the British decipiens at this great length because it has, almost unanimously, been referred to S. fragilis, and considered not, or scarcely, worthy of the rank even of à variety. That it might really be of hybrid origin seems not to have occurred to any botanist *, though several supposed hybrids * Mr. M. S. Bebb, the American salicologist, remarks in a letter to me "1 have always believed that it [i. e. S. decipiens] was a hybrid; but this was 83 e as I got! With the living plant of ¢riandra I am not acquainted. It will not stand the hot sun of our midsuumer months, and can barely be coaxed into & sickly bush, a few feet in height, crowded with dead twigs. S. fragilis, on the other bands in all its forms, fairly luxuriates in our pseudo-Asiatic climate. Now decipiens, sent to me by my dear old friend the Rev. J. E. Leefe as the direct descendant of the plant ot Smith, exhibited, as I now remember (though the fact had for ue significance at the time), a compromise, as it were, between the vigour of and the arrest of growth in inidsummer so marked in the case of triandra.” REVISION OF THE BRITISH WILLOWS. 858 between Salix triandra and S. fragilis have been described. Kerner, for example, distinguished three forms—S. subtriandra, S. alopecuroides, and S. Kovatsii; but of these I have seen authentic specimens of S. alopecuroides only, and the descriptions of none of them fit S. decipiens exactly. If, therefore, it is, as I suppose, a hybrid between fragilis and triandra, it would seem to be one additional to these, which, with yet another, named pro tempore by Andersson S. gracilescens, should all be united under the oldest name, viz. S. decipiens *. There is reason to believe that one of the other forms just mentioned— S. alopecuroides, Tausch—is also British. The Rev E. F. Linton distributed some years ago, under the name of “Salix undulata : Summer-flowering,” a triandrous willow found near St. Neots, Hunts. The specimens I have seen are cer- tainly, in many ways, very like S. wndulata; but at the same time they agree so well—making a little allowance for the abnormal flowering—with Wimmer’s specimens (Coll. 19, Herb. 77) of S. speciosa, Host (S. alopecuroides, Tausch), that I have little hesitation in referring them to that species. Itis desirable, however, that a larger series of normal specimens should be ob- tained. In the locality where Mr. Linton gathered his speci- mens'there is no sign of willow-cultivation, though some species, chiefly S. fragilis, grow there. In Britain S. decipiens is a widely-spread, but not very abundant, Plant. Some botanists think that it is perhaps always planted ; but Smith was of opinion that it was truly wild, though not un- frequently cultivated. In my experience it appears to be as We as its allies S. triandra and S. fragilis, with which it often, tuu ot invariably, grows. Bushes of it occur which have cer- y never been planted, whatever their origin may have been. su. atish illustrations of S. decipiens, Eng. Bot. t. 1937 , and olan A 09. Xxix., represent the g fairly well; but the foliaceous ea (usually absent) at the apex of the petiole seem rather « Me Gs Smith's plate was drawn from a specimen from drawin M 8 garden, May 24, 1808.” Another of the original plant 5 18 marked ‘ Salix decipiens, I think—J. E. S.;" but the ooks to me like fragilis var. b, from the catkins, the adult wares not being shown +. et was written, I have received (through the kindness of Mr. pecuroides, n Weiter of 8. fragilis-triandra (Wimmer's name for S. alo- are nearer $ Sege te in Sweden. Whilst the catkins (d) of these , the leaves &c. are quite those of S. decipiens. TA i inan: who does not know decipiens might, however, easily assumo to be illustrations of S, fragilis. 854 DRE. F. BUCHANAN WHITE'S I have seen specimens of Salix decipiens from the following * Watsonian counties ":— 3, South Devon (Archer Briggs); 6, Somerset (Painter) ; 12, Hants; 17, Surrey (H. C. Watson); 19, North Essex (Leefe) ; 28, Oxford (Druce); 32, Northampton (Druce); 36, Hereford (A. Ley); 88, Warwick (Z. Kirk); 89, Stafford (Fraser); 64, M.W. York (J. G. Baker) ; 65, N.W. York (Ward); 80, Rox- burgh (Brotherston); 81, Berwick (Brotherston); 85, Fife (W. Martin); 88, Mid Perth!; 89, East Perth ! X SALIX SUBDOIA, n. hybr. (S. triandra x S. alba.) This is an equivalent of S. decipiens, with the fragilis element replaced by alba. In many ways it much resembles S. deci- piens; and leaf-specimens might readily be passed over asa form of that hybrid. I have as yet seen the 9 plant only with catkins, but hope that some bushes may prove to be d. The 9 forms a rather low bush with upright branches. Bark of the older twigs pale grey-brown, rather dull, that of the shoots shining yellowish, but becoming more or less brown when dried ; the very youngest shoots occasionally pubescent. Buds lanceo- late, yellowish, with reddish tips when alive. Leaves narrow oblong-lanceolate, tapering at the base, obliquely acuminate at the apex; margin finely and rather closely glandular-serrate, the glands blackish in colour ; upper surface shining pale green, with minute white dots, the veins very slightly impressed ; under surface dull, pale green or more or less glaucous; chief veins slightly raised. Leaves mostly quite glabrous, but the very youngest silky, and the petioles of the older ones and occasionally the underside of the midrib and of the lamina towards the base sometimes pubescent. Catkins about 1 inch long, narrow, cylindrical, dense-flowered, erect or erect-spreading on leafy peduncles about half their own length ; rachis thick, pubescent ; scales pale greenish white, oblong spathu- late, apex truncate-rounded, pubescent at the base, subglabrous 01 the back, inner side concave glabrous, long and densely ciliate or the margin ; capsule small, ovate-conic, with a pedicel a little longe" than, to twice as long as, the small thick yellow inner nectary (no outernectary) ; style almost none, thick, subbifid ; stigmas short, half bifid, recurved-spreading, segments broad, finally brown. The facies of this hybrid is, as mentioned above, much like that of S. decipiens, from which the less shining and greyer bark of the older twigs, the more slender branches, the smaller, narrowe and more finely serrate leaves, the smaller catkins, and sma REVISION OF THE BRITISH WILLOWS. 855 capsules with much shorter pedicels—all of which indicate a con- nection with Salix alba rather than with S. fragilis—serve to distinguish it. Its relation to S. triandra is closer than to the other parent; but the structure of the scales and capsules, as Well as the pubescence of the leaves, when that is present, separate it from that species. The name S. triandra-alba, which might have been applied to this hybrid—if its parentage was beyond doubt,—has already been used by Wimmer for the plant generally known as S. undu- lata. The latter hybrid is, however, now usually admitted to be a hybrid of triandra with viminalis, and not with alba *. I have as yet found S. subdola only on the west bank of the Tay below Perth. X Butz UNDULATA, Ehrh. (S. triandra x S. viminalis.) Though for some time I believed that Wimmer, in thinking that S. alba and not S. viminalis was one of the parents of S. un- dulata, Ehrh. (25. lanceolata, Sm.), was more correct than Andersson and the majority of salicologists, who hold the vimi- nalis theory, I am now persuaded that Andersson’s view is the right one. Wimmer, indeed, at one time considered viminalis to be a more probable parent than alba; but latterly he altered his opinion, though unfortunately he does not, in the ‘Salices’ at least, give his reasons for doing so. But whilst Andersson treats undulata as a distinct hybrid between triandra and viminalis, he curiously gives an equally Independent position to another hybrid (for which he uses Doll's name of multiformis) of the same two species, placing wndulata in the Triandre, and multiformis in the Viminales. Doll more correctly brings all the triandra-viminalis hybrids under one name ; and his example I follow, using, however, the oldest name, undulata, instead of the latest, multiformis. The forms—which have all been described as distinct species— thus brought together are S. undulata, Ehrh., S. lanceolata, Sm., S. Trevirani, Spr., 8, hippophaifolia, Thuill. and S. mollissima, Ehrh. Though Andersson and others adopt the view that S. undulata, Ebrh., and S, lanceolata, Sm., are identical, there is really some oubt as to what Ehrhart’s species truly is. Smith says that * Whilst the dwarf stature and general facies of the bushes incline me still to think that S. triandra and S. alba have both something to do with the parentage M this plant, more recently obtained leaves (from young shoots) strongly recall Jragilis, Tt may be, therefore, possibly a form of S. viridis, though that a 1.5 1o me improbable; or, perhaps, S. decipiens x S. alba (i.e. S. fragilis x ` triandra x S, alba), 856 DR. F. BUCHANAN WHITE'S though it is described as having a pubescent ovary, the original specimens seen by him have glabrous ovaries ; and he suggests that the hairs of the scales have been mistaken for ovarian pubescence. Wimmer thinks that under “ undulata ” several forms have been mixed up; whilst, to add to the confusion, Andersson himself, though citing under Saliz undulata, S. Trevirani, Spreng. and the Sal. Wob. t. 18, repeats these citations under S. multiformis, to which—in the sense in which ‘he uses that name—they really belong. As generally understood, however, undulata, Ehrh., differs from lanceolata, Sm., only in having pubescent instead of gla- brous ovaries, and is said to be extremely rare. From this extreme rarity and from the doubt as to whether the reputed pubescent form was really otherwise identical with lanceolata, Xm., I was inclined to consider the latter a hybrid, as Wimmer declares it to be, of triandra and alba, very similar indeed to the triandra- viminalis hybrids (multiformis, Doll, Anderss.), but distinguished from these by the more distinct and stronger serration of the leaves, which in vernation are revolute and not convolute. Against Wimmer's theory the only point that militated was the fact of the style of lanceolata being more distinct than in either triandra or alba. My doubts on the subject have, however, been removed by the examination of a willow found in Miller’s Dale, Derbyshire, by Messrs. Bailey and Painter. This has abnormal catkins, prO duced at the end of short branches in August. "Whether the ovaries are similar to those of the spring flowers is yet uncertain *; but they differ from those of ordinary lanceolata only in that some of them are more or less pubescent. On the same catkin occur ovaries almost or quite glabrous, some with a little pubes- cence towards the top only, and others more generally pubescent, and with the pedicel also pubescent. Whilst the occurrence of these pubescent ovaries, taken in conjunction with the structure ofthe style, seems to afford tolerably conclusive proof that wn- dulata is a hybrid of triandra with viminalis, it must yet be kept in mind that, judging from what may be seen in other Salices, too much reliance should not be put on the absence or presence of pubescence. Taking, now, the forms which Andersson combines under mul- tiformis, Doll, and Wimmer under S. triandra-viminalis, Wimm., * The Rev. W. Hunt Painter has, since this was written, kindly sent me spring flowering specimens from “this old and large tree.” These have glabrous ovaries. It is desirable to see if the second flowering always shows pubescent capsules. — REVISION OF THE BRITISH WILLOWS. 857 namely a. Salix Trevirani, Spreng., b. S. hippophaifolia, Thuill., and c. S. mollissima, Ehrh., it will be found that they differ from each other chiefly in the amount of pubescence of the various parts, and in the structure of the catkins and capsules, a and b being nearer friandra, and c nearer viminalis. (Here it may be remarked that Andersson’s and Wimmer’s descriptions are not in exact agreement in every particular, nor Wimmer's with his published specimens.) Through the kindness of the Rev. Augustin Ley and Dr. Fraser, I have seen a series of a willow found by them in Here- fordshire and Staffordshire respectively, distributed under the name of “Salix hippophüefolia, Thuill.,” and doubtless the plant meant by the “S. triandra c. Trevirani (Spreng.),” ofthe ‘ London Catalogue.’ * | Whilst Dr. Fraser has found, near Wolverhampton, one d bush only, Mr. Ley reports that the willow found by him is tolerably common on the lower course of the Wye in Hereford- shire, where it is almost uniformly androgynous, though purely male plants do occur. The specimens of the latter which he has sent to me are practically identical with Dr. Fraser's examples, except that perhaps these are a little more pubescent. Comparing them with Wimmer's specimens, I think that, although they do not quite exactly agree, yet they belong rather to a. Trevirani than to b. hippopháefolia. r. Ley's androgynous examples are much more difficult to place; for though, as regards the leaves, they are perhaps nearest Trevirani, in catkin structure they vary a good deal, some being near Trevirani, and others near mollissima, but none of them agreeing with Aippophüefolia as defined by Wimmer. They seem, in fact, to connect Trevirani and mollissima without touching on Aippophüefolia, and show that £riandra-viminalis includes more forms than those described by Wimmer or by Andersson (under multiformis). In addition to the specimens mentioned above, l have seen examples of a willow collected by Mr. A. Brotherston at Carham, in Northumberland, which fits b. hippophdefolia. Con- sidering that both S. triandra and S. viminalis are common British Species, it seems not impossible that hybrids between them may be of more common occurrence than is at present supposed, * It may be mentioned that both Andersson and Wimmer use “ hippophai- folia,” and some other botanists “ hippophdefolia,” to denote Thuillier’s plant r es LFL Par. ed. ii. 514], and that the name Aippophaifolia has been employed for both a and A. 858 DR. F. BUCHANAN WHITE'S though restricted by the difference in the period of flowering of the species. Leaf-specimens of the form Trevirani bear an evi- dent resemblance to those of Saliz rubra; but itis, as mentioned above, with those of the forms undulata and lanceolata that they are more likely to be confounded. Besides the difference in the serration of the leaves, as already stated, distinctions in the female inflorescence are also described ; but these are not all to be relied on. The catkins are sometimes, but by no means always, smaller and slenderer, nor are the scales always darker-coloured. When the capsule is pubescent, that affords a ready mark of distinc- tion from lanceolata; but Wimmer says that it is more usually glabrous though “ punctulato-scabra ;” the pedicel is shorter, and the style often longer and more slender. In the g the sta mens vary from 2 to 3. The form 5. Aippophüefolia has much similarity to a. Tre- virani; but makes a smaller bush, with smaller and narrower leaves, smaller and more slender catkins, and smaller pubescent capsules. The leaves of Trevirani are described by Wimmer as glabrous, and those of hippophdefolia as pubescent when young; but in his examples of Trevirani the young leaves, at least, show pubescence. The form c. mollissima is, in its typical condition, a very dif- ferent-looking plant from either a or b, and might readily be passed over as a form of S. Smithiana (with which, indeed, Smith at first confounded it), from the resemblance of the leaves to some states of that plant, and from the aspect of the catkins. It is, I suppose, on account of this form that Andersson has placed “ multiformis " amongst the Vinimales, and not amongst the Tri- andre. From its Continental distribution the form mollissima might be expected to occur in Britain, but it has apparently not yet been detected. Though I have, in a measure, indicated the distinctions between the various recognized forms of the £riandra-viminalis hybrid, I do not think that in it, more than in other hybrids, should varietal names beretained. Whilst the named forms, as described, seem to have a certain amount of stability, many specimens (including even those published by Wimmer himself) cannot well be placed in any of them, and though there has not yet been found such a complete series—connecting the two parents—as other hybrids afford, this is probably only on account of the rarity ofthe plant. Of all the forms, lanceolata is both the commonest and the least liable to variation; but it is doubtful whether it occurs anywhere in 4 REVISION OF THE BRITISH WILLOWS. 859 truly wild condition, especially since the 9 only is known. Of it I have seen specimens from the following counties, in addition to those recorded in ‘Top. Bot.’ ed. 2 :— 57, Derby (C. Bailey); 77, Lanark (R. McKay); 88, Mid Perth !; 89, East Perth !; 92, South Aberdeen (Trail). Group 2. PENTANDREE. 2. SaLIx PENTANDRA, L. Of S. pentandra, Andersson distinguishes three leaf-forms :— latifolia, with leaves whose length is two or three times the breadth; angustifolia, with leaves three to five times as long as broad and narrower at the base; and microphylla, with thin leaves scarcely more than one inch long. He remarks that latifolia is more usually shrubby, and angustifolia generally arborescent, and thinks that as the latter is the prevalent North-Lapland form, it indicates that the home of the species is in the north (Sal. Lap. p. 15). Walker-Arnott’s experience (and also mine) is that, in a wild state in Britain, S. pentandra is a bushy shrub; but that when cultivated it becomes a tree, with broader and larger leaves than those of the wild plant, in specimens of which from the same marsh they vary much in size and shape. Besides the leaf-forms there is also considerable variation in the size of the catkins ; but in neither case are these characters of sufficient importance to deserve varietal rank. In Britain both the modifications latifolia and angustifolia occur, though many specimens cannot well be referred to one more than the other ; but comparing British examples with Con- tinental—of both of which I have seen a rather large series—it would seem that in Britain there is a greater tendency for the plant to be broader-leafed than in Continental Europe. Indeed, Many specimens from the latter would scarcely, so far as the leaves go, be recognized, at the first glance, as pentandra by a British botanist. In the Linnean Herbarium there are two sheets devoted to S. pentandra. On one, labelled by Linné “3. pentandra,” the d example is, without doubt, pentandra; but the 9 seems a little doubtful. The other sheet bears, in Linné’s writing, “Salix pentandra,” to which Smith has put a “?,” and has also a label, in a now unknown hand, “ Salix pentandra, Flor. Lap. 370. Foliis Subtus cinereis." The specimen is 9, and seems more like S. triandra than anything else. Lastly, it must be noted that Andersson describes the rhachis of 860 ‘ DR. F. BUCHANAN WHITE'S the catkin of Salix pentandra as glabrous ; whereas other authors more correctly allude to it as hairy. It is, of course, possible that Andersson has met with specimens such as he describes. x SALIX CUSPIDATA, Schultz. (S. pentandra x S. fragilis.) In Britain this hybrid has been either overlooked or is very rare, being confined to Shropshire. So far as the leaves go, S. cuspidata, says Wimmer, can scarcely be distinguished from S. pentandra; and this is very evident m many Continental examples, in which, from the leaves alone, it would be impossible to say to what plant they should be referred. The most important distinction between S. cuspidata and S. pentandra lies in the 9 catkins. In the former these are more slender and more tapering, and bear narrower and more cylindrical capsules with longer pedicels. A majority of authors describe the pedicels as three or four times the length of the nectary (as compared with S. pentandra, m which it is at most nearly twice the length of the nectary); but Wimmer says of the pedicel of cuspidata “ brevissimis," aud of pentandra “brevis ;’’ and his specimens show ashorter pedicel in the former than in the latter. I am therefore inclined to think that S. cuspidata, just like other hybrids, shows an instability even in what are generally supposed to be important and constant characters, and that too much de- pendence must not be placed on any one point. Apart from the female flowers, distinctions (all variable) may be found in the leaves (more acuminate, thinner, paler-veined, and sometimes glaucous below, in cuspidata); in the male flowers (fewer-stamened and laxer); in the scales (more hairy); in the size of the tree (bigger); and in the time of flowering (earlier than pentandra and later than fragilis). The form of thestipules is sometimes said to afford an important character; but itis very doubtful if this is the case. Through the kindness of Mr. W. Phillips, F.L.S., I have been able to examine living specimens of the Shropshire plant. Like the British S. pentandra, they have broader leaves than many of the Continental examples. . Though some recent British botanists think that S. cuspidata 18 doubtfully native in its English localities, S. pentandra is admitted to be native as far south as Worcestershire ; and as both it and S. fragilis occur in Shropshire, there seems to be no valid reason, on that ground, why S. cuspidata should not be wild there. In answer to my inquiries, Mr. Phillips has given me the fol- REVISION OF THE BRITISH WILLOWS. 861 lowing information about Salix pentandra and S. cuspidata in Shropshire. S. pentandra seems to be native, making usually a shrub from about 7-10 feet high, but occasionally a treeof about 20 feet. 8. cuspidata (which also appears to be native) attains in its best form a height of 25 feet or more; but another form (more pentandra-like) is only a shrub-like bush. Of the latter I have seen only one plant (which belongs beyond doubt to S. cuspidata); but bushes reported to be similar in appearance grow at comparatively short intervals for twelve miles or more along the Rea Brook. Whether these are all cuspidata or whether, as is probable, some, or most of them, are pentandra, requires investigation. Of the Shropshire plant I have seen the 9 only; and appa- rently the d' has not yet been detected there. X SALIX HEXANDRA, Ehrh. (S. pentandra x S. alba.) , To this rare hybrid I am inclined to refer the following spe- amens ;— l. A plant, in the Edinburgh University Herbarium, collected near Duddingston (Edinburgh) by J. Knapp in 1836, and referred at one time to S. alba, and at another to S. fragilis. This seems most probably a hybrid between pentandra and alba, both of Which grow at Duddingston. The leaves are too young, but, on the whole, are similar to those of authentic specimens of S. hexandra. They are at first clothed with silky pubescence, but become quite glabrous. In Some cases the apex of the petiole is glandular, but not so glan- dular as in Wimmer’s specimens. The catkins (9), with long leafy peduncles, much resemble those of S. pentandra. The capsule is more slender than in that species ; and the pedicel is longer than it is, ag described, in S. hexandra ; but in a hybrid the length Would be subject to variation. The style, almost obsolete, and the short spreading-erect stigmas are like those of S. pentandra. 2. A bush found by me at Restenet, near Forfar, growing with ` pentandra and S. alba. Of this I have not yet seen flowers; but unless to 5S. hexandra, Y do not know where to refer it. Attention may here be called to another probable hybrid of S. Pentandra which I have found near Restenet, where that species abounds. Of this supposed hybrid I have seen leaves only. It somewhat the aspect of S. decipiens; but since S. triandra does not, so far as T know, occur in that neighbourhood, it has probably no Connection with the latter species. The other parent may Possibly be S, phylicifolia, 362 DR. F. BUCHANAN WHITE’S Group 3. FRaAGILES. So much confusion exists in the synonymy of Salix fragilis and its nearest ally, that before entering into a discussion of their forms, it will be first of all necessary to attempt to define clearly some of the essential characters of the three British plants of this group, and then to see what names they should bear. Two of these are usually admitted to be true species, and these I shall term in the mean time A and B. The third is an undoubted hybrid between them, and may be designated A x B. A has the capsule elongate-attenuate from an ovate base, gra dually produced into the style, and hence acute, distinctly pedi- cellate, with the pedicel 2-3 times as long as the nectary. Leaves more or less obliquely acuminate, and, though often at the very first somewhat silky, eventually quite glabrous, and shining above. B has a smaller capsule, ovate-conic in shape, more or less obtuse at the apex, and not tapering into the very short style, scarcely pedicellate, with the pedicels at the very utmost not exceeding the nectary. Leaves usually narrower and smaller, straightly acuminate, more or less silky, rarely eventually sub- glabrous, and slightly shining above. . A x B, in its most intermediate condition, has a capsule larger than B, but smaller than A, conical in shape, more or less obtuse, very shortly styled, and pedicellate, with the pedicel about as long as the nectary. Leaves more or less straightly acuminate, at first somewhat silky, but eventually quite glabrous and shining above, more distinctly serrated than B, but less coarsely than A. But while this is what may be termed the typical state, innu- merable forms, ranging from A to B and showing various com- binations of their characteristics, occur. Some of these are with difficulty distinguished from A or from B, as the case may be. Regarding the name that B should bear, there is no doubt, since all authors are agreed in considering it to be S. alba, L.; but with respect to A and AxB there is much difficulty. Delaying for a moment a consideration of Linnó's description, an examination of other descriptions, figures, and authentic specimens will show that the views of salicologists regarding the names to be assigned to these two species have been widely different. The opinions of the more important writers are 99 follows. (The words of the author indicating which form he had in view are given within brackets and inverted commas.) FS REVISION OF THE BRITISH WILLOWS. 363 A has been called :— Russelliana, Sm., by J. E. Smith in his FI. Brit. 1045; E. B. t. 1808; and Eng. Fl. iv. 186. (Engl. Fl., “germen tapering stalked.") a Russelliana, Sm., by Willdenow, Sp. Pi. iv. 656, n. 7 (“ germini- bus pedicellatis subulatis "). Eusselliana, Sm., by W. J. Hooker in Fl. Scot. 279 (1821) ("germens pedicellate oblongo-subulate"), and in Brit. Fl. 4th ed., 358, n. 14 {“ germens stalked lanceolato-acuminate ”). Eusselliana, Sm., by Forbes, Sal. Wob. 55, t. 28 (“germen tapering stalked,” the plate showing a pedicel three times the length of the nectary). Fragilis, L., by Koch, Syn. Fl. Germ. ed. 1 & 2 (ed. 2, 2. 740, n. 3) (“ capsulis ex ovata basi lanceolatis pedicellatis, pedi- cello nectarium bis terve superante ”). Fragilis, L., by Lindley, Syn. Brit. Fl. ed. 3. 230, n. 3. Russelliana, Sm., by Lindley, 7. c. 230, n. 4. (Lindley follows * the arrangement of Koch; " and though he cites the Eng. Bot. plates for “ fragilis" and ** Russelliana ” respectively, he makes the capsule the same in each, and distinguishes * Eusselliana " by the young leaves being silky and the stipules more acute.) Fragilis, L., by Wimmer, Sal. Eur. 19 (* germina in pedicello revi aut brevissimo, conico-subulata." In Wimmer's specimen, Coll. no. 9, the pedicel is about twice the length of the nectary ). Fragilis, L., by N. J. Andersson, Mon. Sal. 41, n. 28 (^ capsulis Clongato-conicis, attenuatis, pedicello nectarium bis terve super- ante ”), Fragilis, L., by Grenier, Fl. de France, iii. 124 (“ capsule ovoide- conique attenuée au sommet, a pédicelle deux-trois fois plus long que les glandes "1. , Fragilis, L., var. 8, by Walker-Arnott, in Hooker & Arnott’s Brit. Fl. 8th ed., 401, n. 10 (“ ovary lanceolate-acuminate.” A Russelliana, Sm., is given as a synonym of the variety). Russelliana, Sm., by C. C. Babington, Man. Br. Bot. ed. 1l (“ germens stalked lanceolato-acuminate ”’). Fragilis, L., var. 8 Russelliana, Sm., by Babington, l. c. 6th, 7th, and 8th editions (with same description as in the 1st edition, and with, in the last edition, “S viridis, Fr.?,” as a doubtful synonym), Viridis, Fr., by Boswell Syme in Eng. Bot. 3rd ed. t. 1308, as regards the plate, which is a reproduction of Smith’s plate of S. Russelliana, 864 DR. F. BUCHANAN WHITES Fragilis, L., by Boswell Syme, l. c. viii. 205, n. 3, as regards the letterpress (“capsule conical subulate, on a stalk twice or thrice as long as the nectary ”). . Fragilis, L., by J. D. Hooker, Student's Fl., 3rd ed. (“ cap- sule pedicelled,” which definition is rather too brief). A x B, on the other hand, has been called :— Fragilis, L., by Smith, Fl. Brit. 1051; Eng. Bot. t. 1807 ; and Eng. Fl. iv. 185 (* germen ovate abrupt, nearly sessile ”). o. Fragilis, L., by Willdenow, Sp. Pl. iv. 669, n. 51 (“ germinibus subsessilibus lanceolatis ”). Pendula, Ser., by Seringe, ‘Essai, 79. Fragilis, L., by W. J. Hooker, Fl. Scot. 279 (“ germens shortly pedicellate, oblongo-ovate ”), and Brit. Fl. 4th ed. 358, n. 13 (* germens shortly pedicellate oblongo-ovate ”). Fragilis, L., by Forbes, Sal. Wob. 53. t. 27 (“ germen ovate, abrupt, nearly sessile ”). Montana, Forbes, by Forbes, l. c. 37, t. 19 (^ germens nearly sessile, ovate lanceolate ”). Alba, L., y. viridis, by Wahlenberg, FI. Suec. ii. 635 (^ tota glabra viridis." 5S. viridis, Fr., is quoted as a synonym). Fragilis, L., var. y. Russelliana, by Koch, Syn. Fl. Germ. 2nd ed. 741 (capsule not described ; but, from the leaves, must come under A x B). Fragilis-alba, Wimm., by Wimmer, Sal. Europ. 133 (“ germina in pedicello brevissimo conico-cylindracea.” S. Russelliana, Xm., S. viridis, Fr., and S. pendula, Ser., are given as synonyms). _ Viridis, Fr., by N. J. Andersson, Mon. Sal. 43,n. 29 (^ capsulis breve conicis obtusiusculis pedicellatis, pedicello nectarium sub- superante ’’). Fries's own description will be noticed presently. u Fragilis, L., var. pendula, Fr., by Grenier in Fl. de France, ni- 125. (From description of leaves &c., comes here. S. Rus- selliana, Sm., and S. pendula, Ser., are cited as synonyms.) Fragilis, L., var. a, by Walker-Arnott in Brit. Fl. 8th ed. 401, n. 10 (“ovary oblong-ovate "). Fragilis, L., by C. C. Babington, Man. Br. Bot. 1st ed. (“ get mens stalked oblong-ovate ”). Fragilis, var. B. S. fragilis, by Babington, l. c. 6th, 7th, and 8th eds. (“capsule oblong ovate ”). Fragilis, L., by Boswell Syme, E. B. 3rd ed. t. 1306, as regards the plate, which is a reproduction of Smith's plate of S. fragilis. Viridis, Fr., by Boswell Syme, 7. c. viii. 207, n. 4, as regards the REVISION OF THE BRITISH WILLOWS. 365 letterpress (* capsule conical subulate, on a stalk slightly longer than the nectary." Salix Russelliana, Sm., is given asa synonym). Other authors might be cited ; but as I am unable to give the full synonymy from personal examination of the various books (so much confusion exists in the citations even of the best salico- logists, that these cannot be used without having been verified), I abstain from giving more. From the above it will be seen that A has been called :— S. Russelliana, Sm., S. fragilis, L., S. viridis, Fr., and S. fragilis, L., var. Russelliana. And that A x B has been termed ;— S. fragilis, L., S. pendula, Ser., S. montana, Forbes, S fragilis, L., var. Russelliana, S. viridis, Fr., S. fragilis-alba, Wimm., S. fragilis, L., var. pendula, Fr., and S. alba, L., var. viridis. Since thus both A and A x B have been supposed to be S. fragilis, L., Linné’s own descriptions must be referred to. E In the * Flora Lapponica’ (1737), p. 282, No. 849,— which ni described thus, “Salix foliis serratis glabris ovato-lanceolatis acuminatis,"— is often quoted as referring to S. fragilis; but Andersson says that S. fragilis does not grow in Lapland, and that Linné's figure (t. viii. fig. b), which represents a leaf only, must be referred to S. pentandra, which species it is certainly very like. (In Smith’s edition, 1792, there is added, after the dia- gnosis, “Salix fragilis, Sp. Pl. 1443.) In the ‘ Flora Suecica (2nd ed. 1755), p. 847, n. 883, the description runs :— Bam (fragilis) foliis serratis glabris ovato-lanceolatis acuminatis ; pe- tiolis dentato-glandulosis. FI. lapp. 359, t. 8. fig. B. Fi. Suec. 795. Spec. plant. 1017," which is also the description in the ‘Sp. Plant.’ 2nd ed. 1443. In these deseriptions nothing is said about the ovary; and, so far as the leaves go, they might equally well refer to A, A x B, or to S. Dentandra, In fact, the plant of the ‘ Flora Lapponica seems, with little doubt, to be S. pentandra, since Linné says LINN. JOURN.—BOTANY, VOL. XXVII. 2c 366 DR. F. BUCHANAN WHITE'S in his notes that the Lapland plant, which is a lofty bush, differs from the ree which grows in Sweden, which he thus describes :—‘ Salix foliis serratis glabris lanceolatis acuminatis appendiculatis,” and quotes more especially for it Ray's (Hist. 1420), *Salix folio longo splendente, fragilis." In the * Flora Suecica ' he seems to have thought, however, that they were after all the same, as he cites the * Fl. Lap.' aud describes the leaves as ^ ovato-lanceolatis ” instead of simply “ lanceolatis.” It seems, therefore, uncertain whether Linné had in view, so far as his descriptions go, A, or A x B, or both of them. Nor does his Her: barium throw any light on the subject; for the only specimen labelled “fragilis” by Linné has “alba?” added to it by Smith, and seems to be alba d. Hudson and Lightfoot, the immediate English followers of Linné, do not afford any information by which the question can be decided; but Hoffmann (as noticed under S. decipiens) states that “ nomine Sal. fragilis, L., diverse species occurrunt," which, though it does not indicate the exact nature of the Linnean species, is valuable as showing that the subject was in an unsettled state. Not only because he was a Swede, but a botanist of the highest rank, the descriptions of Elias Fries deserve most careful attention. By Wimmer, Andersson, and, in fact, all modern botanists, A X B has been referred to Fries’s S. viridis, with the citation “ Nov. Fl. Suec., Mant. i. p. 43," and “ Nov. Fl. Suec. ed. 2, p. 283.” (As a matter of fact, the name was given earlier, as Fries quotes a prior part—the first edition of the ‘ Novitie ;' and it is mentioned by Wahlenberg, * Flora Suecica,’ 1826, under S. alba, with the citation '*S. viridis, Fr. Nov. p. 120,” while the dates of the works mentioned above are 1832 and 1828 respectively.) As the ‘ Mantissa’ seems to be considered the most important citation, we will first of all examine it. The part relating to the Willows is entitled * Commentatio de Salicibus Suecie." Here the species and forms we are now discussing are mentioned as follows :—3. S. fragilis, with varieties B. S. pendula and y. S. vi- tellina ; 4. S. viridis ; and 5. S. alba. S. fragilis is described as having the later leaves " subsericeis," and the capsules “ subsessilibus ovato-conicis.” It is referred to the S. fragilis of Linné (“in Itin. Scan. p. 200 stabilita ") and of Sinith. The branches are stated to arise at right angles to the trunk. REVISION OF THE BRITISH WILLOWS. 367 The var. B. Salix pendula differs by its more elongate, slender, at length pendulous branches and smaller capsules. Linné's ‘Flora Sueciea, first edition, No. 812, is cited under it, and also S. Russelliana, Sm. “ex spec." The leaves are said to be much narrower, the later ones often notably silky, aud the capsules more evidently pedicellate. In both forms the catkins are de- scribed as making right angles with the branch, aud the ripe 9 catkins as pendulous. S. viridis is said to have very glabrous leaves, and the capsules to be “ pedicellatis ovato-subulatis." The catkins are afterwards described as erect, and the later leaves pilose below. Fries adds that it agrees well enough with S. decipiens among the Smithian species. In the second edition of the ‘Novitie’ the capsules of s. viridis are described as “ subpedicellatis ovato-subulatis” and as with “ pedicello brevissimo ;” and the angles made by the branches are alluded to, viz. 90? in S. fragilis, 60? in S. viridis, and 35° in S. alba. As, however, this work is four years earlier than the ‘Mantissa,’ the description in the latter must be consi- dered as superseding that in the former work. In a later work, ‘ Flora Scanica,’ 1835, S. fragilis is distin- guished by the later leaves being silky and catkins pendulous. S. viridis by very glabrous leaves and erect catkins. So far, then, as Fries’s descriptions go, his S. fragilis, since it has subsessile and ovate-conic capsules, must belong to A x B, and his S. viridis, from its pedicellate and ovate-subulate cap- sules, to A. The leaves also indicate the same conclusion; but the description of the direction of the branches suggests the reverse. Fries's published specimens (in his ‘Herbarium Nor- male’), so far as I have seen them, belong—S. fragilis to A, and S. viridis to AxB. From this confusion the chief deduction seems to be that Fries’s views were not always the same; and that, as regards at least the essential part of the descriptions in the “ Commentatio," his S. fragilis is the same as Smith's. A result, therefore, whieh we appear to have arrived at is is :— lst. That there is no certainty, but absolute uncertainty, re- garding the species which Linné himself had in view. 2nd. That while Smith, Willdenow, &c. (and Fries in his de- scriptions) have considered AxB to be S. fragilis, L., Koch, mmer, Andersson, &e, have taken A to be that species. 2c2 868 DR. F. BUCHANAN WHITE'S The question is, therefore, which of these two parties is right, —a question which, it seems to me, it is almost, if not quite, 1m- possible to decide. I am inclined to think that it is not unlikely that A x B (the Smithian Salix fragilis) has at least as much claim to be considered Linné’s S. fragilis as A has. If this claim could be proved, then A would = S. Russelliana, Sm., and AxB would — 8. fragilis, L. ; but as there is so much uncertainty, and as an adoption of Smith's views would entail great synonymic changes, it seems more expedient to follow Koch, &c., and to con- sider A as being the Linnean S. fragilis, in which case A X B= S. fragilis x S. alba. 8. SALIX FRAGILIS, L. S. fragilis, as it has just been defined, namely, with S. viridis as well as S. decipiens eliminated from it, is not, as it occurs in Britain, a species subject to any great range of variation. The points of distinction between it and S. decipiens have already been pointed out, and those between it and S. viridis may be better considered when treating of that plant. Though in Britain, as in Continental Europe, there is no great difficulty in distinguishing it from S. alba, yet it is so closely allied to the latter, that, as Andersson remarks, in some regions of Asia they are united by so many forms, that it is not easy to point out the difference between them. For this reason, he thinks that pos- sibly it originated there, and has thence immigrated into Europe. If this be the case, it must have occurred at a very early period, if Saporta is right in his determination of certain plant-remaivs of the Pleistocene tufa of the Seine valley, not to mention Heer’s discoveries in Swiss Miocene deposits. Be that as it may, S. fragilis is, at the present day, in Britain as native, to all appear- ance, as most other species of lowland willows. As mentioned above, S. fragilis does not in Britain show avy great range of variation; but an examination of any extensive series and a comparison with Continental specimens will show that there are two tolerably distinct forms which seem worthy to rank as varieties. These may be called a. genuina and f. britannica ; and the distinctions between them lie chiefly, if not entirely, in the flowers. Ina the d catkins are rather dense-flowered, with the stamens conspicuously longer than the scales, whilst in d the catkins are lax-flowered, and the stamens are scarcely longer than the scales. Hence in e the stamens aud in 6 the scales form the conspicuous feature of the catkins. In the 9 the differences REVISION OF THE BRITISH WILLOWS. 369 between a and () are not so striking, but in good specimens are sufficiently evident. In a the ovary is much wider at the base than in 8; and may be described as ovate-lanceolate, that of B being lanceolate-subulate. The scales, though variable, are perhaps more usually shorter in a than in 8; the catkins, espe- cially when young, rather denser-flowered and stouter; and the styles and stigmas rather thicker. The chief difference, however, is in the shape of the ovary. The two varieties show no constant differences in leaf-charac- ters, though I am inclined to suspect that a is generally broader- leafed than f. Whilst both forms occur in Britain—, however, being very much the more abundant one—a,so far as I have seen (from specimens and figures), is the only variety found in Continental Europe. In Britain a seems to be decidedly rare. Ihave found one d plant in Perthshire; Mr. A. Brotherston has gathered both sexes in Roxburghshire; and a d collected by Mr. T. R. Archer Briggs in South Devon seems also referable to it. Since the 9 is Smith's Saliz Russelliana (of which the 9 only was known to him), perhaps 8 ought to retain that name as a varietal designation ; but as the name “Russelliana ” has been the subject of so much confusion, the propriety of keeping it up seems doubtful. Moreover, it is possible that Forbes’s S. mons- peliensis is the d of 6. Specimens received from Kew Gardens with the name S. monspeliensis belong to the var. ; but Forbes's figure (Sal. Wob. t. 30) is doubtful, as, though the single flower Shows the long scale of (3, the rest of the figure is more like a. The E. B. t. 1808, and Sal. Wob. t. 28, of S. Russelliana re- present B; but in both figures the ovary is a little too obtuse at the apex—in the original drawing for the Eng. Bot. figure it is not obtuse. The g catkins figured in the plates of “fragilis” in Eng. Bot. t. 1807, and Sal. Wob. t. 27, suggest f ratner than a, as do the single g flowers of the former, while in the latter they are like those of a. Itis possible, however, that in both cases they have been taken from S. viridis. With the original drawing of the d for Eng. Bot. a single catkin is preserved, which, if not viridis, is the var. a of Sragilis. Several leaf-varieties of S, fragilis have been described, founded on the breadth of the leaf, or on the colour of the underside ; but "ey are not of sufficient distinctness or importance to be worthy of retention, Finally, since in Britain S. fragilis has been so much confused 870 DR. F. BUCHANAN WHITE'S with Saliz viridis and S. decipiens, and since it is possible that (as in several Continental localities) it may be in some places rarer than S. viridis, it will be necessary to work out afresh its distribution, though it probably occurs throughout. I have specimens from the following counties :— 8, S. Devon (Archer Briggs); 16, W. Kent (E. Edwards); 17, Surrey (E. S. Marshall) ; 19, N. Essex (Leefe); 20, Herts (B. Daydon Jackson); 23, Oxford (G. C. Druce); 27 or 28, Norfolk ; 32, Northampton (Druce); 34, W. Gloucester (J. M. White); 36, Hereford (4. Ley) ; 37, Worcester (E. F. Towndrow) ; 38, War- wick (T. Kirk); 40, Salop (Leighton); 47, Montgomery (Eyre Parker) ; 57, Derby (C. Bailey) ; 64, M.W. York (J. G. Baker); 65, N.W. York (Ward); 66, Durham (Middleton); 80, Rox- burgh (4. Brotherston) ; 88, Edinburgh (Maughan) ; 85, Fife l; 87, W. Perth!; 88 Mid Perth!; 89, E. Perth!; 90, S. Aberdeen (Trail). 4. ÑALIX ALBA, L. Although Andersson describes eleven varieties or forms (the majority, however, not being natives of Europe), and Wimmer four principal forms, S. alba does not in Britain present any great range of variation, with the exception of the remarkable modifi- cation S. vitellina. Most of the British specimens which I have seen are in the direction of S. cerulea, Sm., a form which in structural differences can be separated from typical S.alba only by its adult leaves being less pubescent. In other respects S. ccerulea,—probably the “Huntingdon Willow” of arboriculturists, though Smith calls “Russelliana” the Bedford or Huntingdon Willow,—is said to differ from ordinary S. alba by its more rapid growth and the greater value of its timber, and hence would be selected for planting. Thus all our trees being probably self-sown escapes when not actually planted, would most naturally belong to this form. There seems thus to be no good grounds for retaining cerulea as a variety. In its extreme form the adult leaves are nearly gla- brous; but between this state and the very pubescent form argentea, Wimm. (S. splendens, Bray), all gradations occur. Of the extremely pubescent form, I have seen no British examples. As regards leaf-form, there is some degree of variation in British specimens, but not so sufficiently marked as to deserve distinction. Lecfe bas noticed (in Sal. Exs.) that the leaves of REVISION OF THE BRITISH WILLOWs. 371 the 9 are generally broader; and, so far as I have seen, they are constantly broader in proportion to their length, and hence possibly belong to Wimmer’s form ovalis. Salix vitellina, L., is reduced by Koch, Wimmer, Andersson, &c. to the rank of a mere form of S. alba, distinguished only by the colour of its bark and its generally narrower leaves, and said to be often produced either by being cut over every year, or by a diseased condition of the plant. Whether it be a diseased condition or not, it certainly does not depend for its characters on annual lopping; and, in addition to the colour of the bark, has several important structural differences to separate it from ordinary alba. The catkins are much more slender and proportionately, if not actually, longer, with usually more scattered flowers, and remark- ably long and narrow scales. The leaves are smaller and usually of a paler or more yellow-green, and frequently less pubescent. Boswell Syme thinks that the ovaries are always abortive; but Whilst I have not had sufficient opportunities of studying the plant to be positive, I am inclined to think that he is wrong on this point. On these grounds, I consider that vitellina is worthy of at least varietal rank. Fries thought that “ S. vitellina, L.," pertains to S. fragilis. In Linné’s herbarium are two specimens labelled by Linné “ 5. vitellina.” To the d Smith has put a “?” and to the 9 “alba?” The d has the long scales of vitellina; the 9 has short scales, but long narrow catkins ; the bark in both is now purple-brown, While the leaves are too young to afford much information. On the whole, I think these specimens are nearer vitellina than typical alba. Whether S. alba is native in Britain is perhaps doubtful, for, whilst usually planted, yet self-sown plants occur. The var. vitellina is said to occur chiefly in osier-grounds, but in Norfolk Crowe (Smith's « Engl. Fl.) thought it was “ truly wild.” "Ban vinipis, Fr. (S. fragilis x S. alba.) 5 viridis, Fr., may be taken as the central or typical form of a illow which presents a regular series of gradations between S. frag ilis on the one side and S. alba on the other. Consequently ìt 15 considered to be a hybrid between these two species; but if, as Andersson says (as has been already noticed under S. fragilis), these two species are so connected in some parts of Asia that it ^s difficult to separate the one from the other, it seems quite pos- 372 DR. F. BUCHANAN WHITE'S sible that the plants known as Salix fragilis and S. alba may be only extreme states of one species, and that S. viridis is the inter- mediate condition. In favour of this view may be cited Hartig's statement that in N.W. Germany, where S. fragilis is very rare, S. viridis is very abundant, and. Kerner's that in Lower Austria it is scarcely less common than S. fragilis itself. The question of the specific identity of S. fragilis and S. alba is one, however, that must be studied in the supposed metropolis of the species in Asia, where it can alone be determined whether the connecting forms mentioned by Andersson are inter- mediates or hybrids. In Western Europe, into which they have immigrated or been brought, they must, in the meantime, be treated as distinct species and S. viridis as a hybrid between them. In its most intermediate or typical condition S. viridis (or at least the 9 plant) is not diffieult to recognize, but those states (especially of the g ) which approach one or other of the parents are by no means easy of determination. Compared with S. fragilis, typical S. viridis has leaves of à darker green *, rather less shining, more finely serrated on the margin, and less oblique towards the apex ; the d catkins rather dense-flowered, narrower and longer; the 9 catkins more(usually ) erect and more slender; capsules smaller, paler when dried, conical-cylindrical, contracted into the style, and hence more or less obtuse at the apex (instead of lanceolate-subulate or ovate- lanceolate, attenuate into the style), and with a shorter pedicel not exceeding twice the length of the nectary (instead of 2-3 times as long t). From S. alba, S. viridis differs by the broader and larger leaves, very soon quite glabrous, and more shining above; longer and less dense-flowered d catkins ; and larger, usually more distinetly pedicelled and distinctly styled capsules. The capsules hold an intermediate place in size between those of fragilis and of alba. Taking the average of fragilis capsules as 7 mm. in length, and of alba as 2 mm., those of viridis are about 5 mm. . The habit of the tree is also different from that of both its * Tn his directions to the artist, noted on the original drawing for Eng. Bot., Smith says, regarding Russelliana, “ green of every part lighter than in S. fragilis (i. e. his fragilis, which is viridis), and also “the midrib is much broader in this than in fragilis.? ‘ + The pedicel of fragilis, whilst usually 2-3 times the length of the nectary, is sometimes barely twice the length, but varies in the same catkin. REVISION OF THE BRITISH WILLOWS. 373 — parents. In Salix fragilis the branches (or at least the main ones) make with the stem an angle of 90°, well shown in the lower branches of the figure of * Johnson's Willow” (referred to S. Russelliana, Sm.) in the frontispiece of Salictum Woburnense. In alba the angle is 35°, and in viridis it is said to be 60°. I rather think (from measurements I have made) that the angles made by the twigs show a somewhat similar difference—those of fragilis forming the widest and of alba the narrowest angles, viridis being more or less intermediate. Too much dependence, however, must not be placed on this character. Now whilst the characteristics of the typical viridis are as mentioned above, the hybrid, as met with, more frequently shows a departure from these towards either fragilis or alba, till finally it is almost impossible to separate it from one or other of these species. It may also have one set of organs very similar to one of its parents, and another set resembling those of the other parent. Thus the leaves and twigs may be almost inseparable from those of alba, whilst the capsules may be almost identical with those of Sragilis, or vice versá. It is hence almost or quite impossible to frame a short description which will include all the characters of this variable plant. In practice, however, the trained eye of a sali- cologist will detect differences which cannot easily be expressed in Words, S. viridis is widely distributed in Britain, occurring from Cornwall and Surrey to Perth. How it has hitherto escaped re- cognition is rather surprising, but is probably due to the fact of the failure of salicologists (both British and Continental) to dis- Cover that the Sragilis of Continental botanists and Smith's fragilis Were different plants, and the consequent confusion regarding, and erroneous determinations of, Smith’s Russelliana. Andersson, moreover, in some quite unaccountable way, not only failed to recognize British specimens of viridis, but actually named them Sragilis,” and thus led British botanists (who would place a deserved confidence in the opinion of that great salicologist) to naturally believe that viridis did not oceur in Britain. Not to Specify other instances (e. g. several examples in Kew herbarium *), the case of the specimens in Leefe's * Salictum Britannicum Exs.’ may be mentioned, “ No. 52. S. fragilis, E. B. t. 1807,” of which , * One instance is very curious. The more mature 9 example on a sheet of Pecimens from Upsala, collected and named “ fragilis” by Andersson, must, vinis S to his description in the ‘ Monographia,’ belong, beyond doubt, to , 374 DR. F. BUCHANAN WHITES D Leefe says, “ My plant agrees well with the figure of E. Bot., | was examined in Watson's herbarium by Andersson, who m D this note regarding it (Watson in ‘ Botanical Gazette, No. ` 1851) :—" Specimina foliifera ad S. viridem, Fr.—amentifera " S. fragilem, L., pertinent." * I have not seen the specimens : Watson's herbarium, but there seems no reason to suppose tha they are different from other examples sent out under the same name and number by Leefe at the same time, and similar also : specimens published in his second fasciculus. These post are, beyond doubt, the Salix fragilis of Smith, and have ana P 1 sessile, ovate-conic, obtuse, short-styled ovary. Andersson 14 not at that time written the ‘ Monographia;’ but in it he an under S. fragilis, L., the figure in Eng. Bot. t. 1807 , which, as is t ° case with Leefe’s specimens, shows characters in direct opposita i to his diagnosis of fragilis, but quite in accordance with the e- scription of viridis. What adds to the unintelligibility of An T son's note on Leefe's specimens is that the leaves of the latter do not belong to the most intermediate condition of viridis, but ` somewhat in the direction of fragilis, and hence show a E knowledge of viridis on Andersson's part. An explanation 9 the mystery may be this, that, in the first place, there was time only for a hurried examination of these and the Kew specimens and, in the second, a belief, adopted from others, and not verifie by personal examination of his works, that Smith's fragilis was the same as the continental plant. un Both Wimmer and Andersson recognize the fact that S. viridis varies in the direction of one or other of its parents, but they do not describe so great a range of variation as actually exists. The modifications described by Andersson are a. fragilior (in the dir ec- tion of fragilis), B. excelsior (the typical or most intermediate condition), ard y. albescens (near S. alba). Wimmer has also three forms—a. viridis (with leaves green below), b. glabra (leaves glaucous below), both being glabrous t ; and c. vestita, which 18 the same as Andersson’s albescens. . Nyman (‘ Conspectus’) thinks that three varieties are of major value, viz. S. Russelliana, Sm., S. excelsior, Host, and 8. DÉI lustris, Host. — Russelliana, Sm., is, as has been shown, not rightly * Andersson (Mon.) cites Leefe's No. 50 (decipiens) under fragilis, but does not mention Nos. 51 and 52. . f t He says also “ gemmis glabris," yet the autumn buds of his exemplars © glabra are pubescent, but seem to become glabrous by spring. REVISION OF THE BRITISH WILLOWS. 375 placed here, even as a synonym ; ewcelsior is, according to Wim- mer, the same as his var. glabra, and palustris his var. vestita ; but Andersson’s determination of Host’s species is different. These varieties, whatever names be adopted for them, are marked by too indefinite and inconstant characters to be of much practical value. On the whole Andersson’s varietal names fragilior and albescens, if taken in the sense that they indicate a departure from the central form in the direction of fragilis or of alba, are to be preferred ; but, as mentioned above, the vegetative organs may show affinity with one, and the reproductive organs with the other, of the parents. If these names are used at all it should be with reference to leaf-structure only. In Britain Salix viridis exhibits a considerable range of variety. It may be of some interest and utility to briefly notice some of the forms which 1 have seen. From Worcestershire, Mr. R. F. Towndrow kindly sent me living specimens from seven trees. Two of these are so closely related to S. alba that if they had not formed part of the series received, I should have called them rather untypical S. albu. Two others are also on the alba side of the type or excelsior, which they connect with albescens—one being nearer the latter variety than is the other. Two others exhibit relationship to the other parent, one having leaves scarcely distinguishable from those of S. f ragilis, and the other being much nearer the type. The re- maining tree seems nearest of all to typical S. viridis and has leaves quite different from any of the others, and resembling those of t. 27 in Sal. Wob. and not unlike Eng. Bot. t. 1807. The cat- kins (3) have the filaments apparently shorter, in proportion to the length of the scales, than in Swedish examples of S. viridis, Suggesting a derivation from the var. britannica of S. fragilis. The broad dark green leaves recall those of S. alba 9. Of the same leaf-form as the last is a 9 plant collected in Oxfordshire by Mr. G. C. Druce. It has long and slender catkins, and capsules resembling those of S. alba, but shortly pedicelled. Very different from these specimens are a series of examples collected by Mr. T. R. Archer Briggs in South Devon and East Cornwall. They are all g and, unfortunately, most of them have no adult leaves, Hence their affinity is not so easy to determine, but out of eight trees seven seem to be related to S. alba rather than to S. fragilis. Three of them are remarkable for having more or less partially triandrous, and evea tetrandrous flowers, a 376 MR. F. BUCHANAN WHITE'S condition which, though probably of rare, is not of unknown occurrence. Whilst in most of Mr. Briggs’s specimens the twigs are rather slender and reddish in colour, the catkins vary both in size and direction. Some of the specimens are rather like No. 1955 of Billot’s ‘ Exsiccata, published as Salix fragilis, but which agrees with the description in Grenier and Godron’s ‘ Flore de France’ of S. fragilis var. pendula (S. pendula, Ser.), which Wimmer says is a form of his var. vestita of S. viridis. In Roxburghshire, Mr. A. Brotherston has found a few trees which may be described as having the leaves of S. alba, and the capsules, except that the pedicel is short, of S. fragilis. He also finds what seems to be the d of the same, and I have seen 4 similar plant from Brandon, Warwickshire (Z. Kirk). These would be called var. albescens. In Perthshire, S. viridis occurs in several places on the banks of the Tay, along with S. fragilis and S. alba. In one place the trees have evidently been planted, but have almost certainly been brought from some other and adjacent part of the banks. They are all 9 and, whilst distinctly S. viridis, closely approach S. fragilis in character. In another locality a single tree grows. This is much older than those in the above-mentioned station, and is, in all probability, self-sown. It is in many ways like S. alba, from which, however, the more glabrous leaves and distinetly though shortly stalked capsules at once distinguish it. From some other stations I have as yet seen leaf-examples only. From the erroneous conceptions of Smith's Russelliana by many salicologists, it is not to be wondered at that the citations of descriptions and figures, as regards both S. viridis and S. fra- gilis, are in many cases wrong. Andersson, for example, cites Sal. Wob. t. 27, and Wimmer both that and t. 29 and t. 30 under S. fragilis, L. ; but t. 27 is viridis, t. 29 is decipiens, and t. 30 is fragilis. Under S. Jragilis-alba (=viridis, Fr.) Wimmer refers to Sal. Wob. t. 98 (Russelliana), which represents fragilis, and to t. 19 (montana), which, from the description “ germens nearly sessile," as well as from the figure, is viridis. The references to Smith’s descriptions and figures, and to those writers who have followed Smith, are likewise all wrong, though, while giving * Russelliana, Sm.," as a synonym of the fragilis-alba hybrid, both Wimmer and Andersson confessed that in their opinion the identity of Russelliana, Sm., was altogether doubtful. It is evident, therefore, that the synonymy and citations—both British REVISION OF THE BRITISH WILLOWS. 377 and Continental—ot Salix fragilis and S. viridis require careful revision, since many of the existing citations have not been, it would seem, verified by personal examination of the works cited. Both Wimmer and Andersson appear to be of opinion that viridis rarely, if ever, occurs otherwise than as an introduced plant. Whether this be the case in Britain, further investigations are required to show, but it is probable that it occurs sponta- neously (7. e. self-sown) and as wild, but not more so, as either of its parents. From the nature of some of the forms it is likely that they have arisen from variations in the mode of cross-fertili- zation, some having sprung from fragilis g x albaQ or vice versá, and others from the crossing of viridis with one of ite parents. Differences, too, have probably originated through one or other of the two varieties of fragilis being concerned in the parentage. Experiments like those conducted by Wichura are necessary to decide these questions. S. viridis has been found in the following counties :—2, E. Cornwall( Archer Briggs); 3,8. Devon( Archer Briggs); 13, Surrey (Winch, A. Bennett) ; 19, N. Essex (Leefe) ; 23, Oxford (G. C. Druce) ; 29, Cambridge (J. Holme); 30, Bedford (C. Abbot); 37, Worcester (R. F. Towndrow) ; 38, Warwick (T. Kirk) ; 80, Rox- burgh ( 4. Brotherston); 83, Edinburgh (Boswell Syme); 88, Mid Perth! Possibly also in Stafford and Derby. B. DIANDRZ. Group 4. CaPRES. This group consists of a number of very closely allied species, three of which have been found in Britain. These three—$. cmerea, L., S. aurita, L., and S. Caprea, L.—are the most widely distributed members of the group, occurring throughout Europe at least. Whether they are really distinct species is a disputed Point, but salicologists on the whole are tolerably unanimous in retaining them as such, although it is not to be denied that, either thata mediate or hybrid forms, they are so closely connected ine n 18 difficult to point out distinctions which will hold good in Ad case. Wimmer, however, is of opinion that when studied ano wing state they may be separated without much trouble— Pinion for which there seems to be justification, but which ?** not apply to the determination of dried, and consequently 378 DR. F. BUCHANAN WHITE'S often imperfect specimens. Such examples it is often im possible to place. Whilst the older British botanists, Smith and his followers, thought that there were in Britain five gpecies— the three men- tioned, with the addition of Salix aquatica, Sm., and S. oleifolia, Sm.—-the author of the latest British Flora(‘The Student's Flora’) has, on account of some remarks of Andersson, reduced S. cinerea to the rank of a subspecies of S. Caprea, and suggests that S. aurita is probably only another form. But, though in continental Europe S. cinerea is not rarely, as Andersson says, to be distin- guished without difficulty from some forms of S. Caprea, in Britain this does not seem to be usually the case, since our common form of S. cinerea is much more closely related to S. aurita than to S. Caprea. From the intimate alliance between these three species, and as they flower much about the same time, they, as might be expected, readily cross with each other, and to identify the parentage of the three hybrids thus produced is often most difficult. 5. SALIX CINEREA, L. If a series of Continental examples of S. cinerea be compared with a series of British specimens, it will be seen that, though, perhaps, some of the examples may be tolerably similar, there is, on the whole, an absence of exact identity between the series. Moreover, if the descriptions of the species by the Continental salicologists be studied, it will be found that they do not, in some particulars, quite fit the British plant. That this appears to be a matter of some importance may be assumed if we consider that several species, of more or less restricted distribution, have arisen either from S. cinerea, or from its possible progenitor S. Capret; or from the same stock. Thus, peculiar to alpine and subalpine (especially limestone) regions from France to Transylvania there is S. grandifolia, Ser., a species which has been confounded with S.Caprea and S. cinerea, and has even been supposed to be an alpine modification of the latter. North of the range of this species, and chiefiy in alpine and mountainous districts of Silesia, occurs S. silesiaca, W., à species which has very great affinity and similitude to S. cinerea, S. Caprea, and S. aurita. Of this a subspecies is found in the Caucasus. In the Mediterranean region occurs S. pedicellata, REVISION OF THE BRITISH WILLOWS. 379 Desf., which is connected with Salix silesiaca by S. grandifolia, and has so much affinity with these that it might be considered to be asouthern modification of one of them. S. pedicellata forms the connecting link with S. canariensis, C. Sm., a species very similar to it, restricted to some of the North-Atlantic islands. Besides these, certain subspecies of S. cinerea occur in Russia and in Persia. l From these instances it would seem that there is a tendency in S. cinerea aud S. Caprea to develop local races, some of which are so sufficiently distinct as to be considered as species. Hence it would not be very surprising if in Britain an insular form should oceur, and the differences between it and the Continental plant become greater by separation, since, on account of the absence of economie value in the species and its great abundance, living plants from the Continent would not likely be commonly, if at all, breught to this country. l In such closely allied Willows as the Capree the characteristic distinctions are more easily seen than described. It is thus rather dificult to put in words the differences between the Conti- nental and the British cinerea, though the facies is sufficiently noticeable. In the first place, the European plant has, as the name implies, an ashy-grey appearance, which is, in great measure, absent from the British form. This ashy-grey colour is owing to the pubes- cence of the twigs (“ veluti incani," Wimmer; *griseo-tomentosi ” and “ incani," Andersson ; cano-tomentosi,? Koch; “ grisâtre- tomenteux,” Grenier) and of the underside of the leaves (^ grisea,” ` cinerea,” * de couleur cendrée "), and sometimes of the upper surface also. The British plant is, as regards the great majority ot specimens, much less pubescent. The pubescence of the twigs 5, Moreover, not of a hoary-grey colour (except, perhaps, in the youngest shoots), but inclining rather to fuscous black; and the Pubescence of the underside of the leaves has most usually a more or less considerable admixture of shining, rust-coloured, short, crisped hairs, which, with scarcely an exception (one will be noted), none of the Continental specimens show nor do the Conti- Dental descriptions mention. In the second place, the average size of the leaves seems to be 8s in the British than in the Continental plant. From this difference in the nature of the pubescence and in the ze of the leaves, it follows that the British S. cinerea is more remote from S, Caprea than the Continental form, and rather nearer to S, aurita. le si 380 DR. F. BUCHANAN WHITE'S One of the European examples of Salix cinerea makes a nearer approach to the British form than any of the others. This belongs to the modification which occurs in Portugal, and which has been described as a species under the name S. atrocinerea, Brot., which Andersson treats as a synonym of cinerea. In the only specimen of this form which I have examined, the pubescence of the leaves (which, though thinner, are not unlike those of the British plant) is scanty and has a mixture of ferruginous hairs, but the catkins (8) are remarkably pubescent and unlike ours. If one can judge from this single specimen, S. atrocinerea is worthy of more consideration than Andersson has given it. In the catkins I have not yet detected any constant difference between the British and Continental S. cinerea. Andersson quotes an observation of Lange that in cinerea the d' catkins are centri- fugal and in Caprea centripetal, and Hoffmann’s illustrations of cinerea seem to show this. According, however, to my observa- tions our cinerea has most usually centripetal inflorescence. Without, however, a comparative study of living specimens of the Ccntinental cinerea, I am unwilling to ascribe to the British form varietal or subspecific rank, though it may be tbat further investigations will show that it is worthy of such. Wimmer describes one variety only (B. spuria, Wimm.) of 5. cinerea. This differs solely by its narrow lanceolate leaves, and forms very like his specimens (Coll. No. 32) occur in Perthshire and elsewhere. Andersson notices three leaf-varieties—a. lati- Jolia, P. longifolia (which includes spuria, Wimm.), and y. brevi- Jolia, with a subvariety microphylla resembling S. aurita. He has also a catkin-variety (6. laxiflora), flowering later and with a leafy peduncle to the catkin. "These leaf-varieties are useful only as indicating a considerable range of variation in the shape of the leaves, since there are all degrees of gradation between them. As has been mentioned above, British cinerea seems, on the whole, to be less pubescent than the Continental plant. Jn anot inconsiderable number of specimens the pubescence of the year- old twigs, usually stated to be an important characteristic of the species, is almost or quite absent. Whether these specimens are true cinerea or hybrids it is impossible to say, since in their other characters they do not depart from cinerea. Frequently, too, the adult leaves are nearly glabrous. Smith described, as has been already said, two supposed species which more modern botanists have placed as varieties or synonyms REVISION OF THE BRITISH WILLOWS. 381 of Salix cinerea. Boswell-Syme and, following him, J. D. Hooker term them and cinerea genuina slight varieties, which so run into eachother that it is often impossible to refer a specimen toone more than to another, the distinctions being that oleifolia has narrower leaves with (as in cinerea genuina) reddish-brown hairs beneath, and aquatica obovate thinner leaves with usually white hairs beneath. Andersson considers them as synonyms of S. cinerea, aquatica being most probably his var. latifolia. Wimmer makes S. oleifolia a synonym, but, on account of Sal. Wob. t. 127 and Dóll's opinion of Eng. Bot. t. 1437, refers aquatica to the hybrid S. Caprea- cinerea, Wimm. Walker-Arnott cannot distinguish them as well- marked varieties, and points out the fallacy of eharacters derived from the stipules, in whose structure Smith placed reliance. W. J. Hooker retains them as species, but states Borrer’s opinion that their characters are unsatisfactory. These quotations will be sufficient to show that the general opinion is that S. aquatica and S. oleifolia are scarcely distinct, as varieties even, from S. cinerea, an exception being Wimmer's idea (derived from figures only) that S. aquatica is a hybrid between S. cinerea and S. Caprea. One or two points in Smith's descriptions are not alluded to by modern botanists. S. cinerea he describes as a tree 20-30 feet high if left to its natural growth. S. oleifolia is also a tree; but S. aquatica is generally bushy, rarely forming a tree. The catkins of S. oleifolia are as large as those of S. Caprea, those of S. aquatica being much smaller and more like those of S. cinerea. He also says that S. aquatica is “ most related” to aurita, and places the species In this order. 8. cinerea, S. aurita, S. aquatica, S. oleifolia. Iu Smith's herbarium specimens of both S. aquatica and S. olei- Jolia, from “ Mr. Crowe's garden," are preserved. These are in- structive in several ways, and show some discrepancies with the descriptions, The specimen of S. aquatica much resembles „aurita in twigs and catkins; the leaves are thin, not very hairy, Nae wë above, margins slightly undulate-crenate, and und er- with eit reddish-brown hairs. It looks like a hybrid of S. aurita eaves s er S. cinerea or S. Caprea, the pubescence of the young slender ene the latter. The S. oleifolia specimen has rather leaves aurita-like twigs which are not very pubescent; the entire resemble those of cinerea, have slightly revolute and sub- catkins argıns, and the underside has reddish-brown hairs; the or eine (3) are larger thzn those of the usual aurita, but small EM and their scales are suggestive of aurita. With the : JOURN.— BOTANY, VOL. XXVII. 2D 382 DR. F. BUCHANAN WHITE'S drawing for Eng. Bot. t. 1402, Salix oleifolia, a leaf is preserved. This, while rusty on the veins beneath, is covered below with coarse white woolly pubescence, and bas the upper surface coarsely hairy; margins crenate-serrate and slightly revolute. The specimens published by Leefe are interesting, not only as ilustrating the opinion of the British botanists who succeeded Smith, but also Andersson's earlier ideas; for at the time that he examined Leefe's examples the hybrid origin of many willows had not been recognized. Leefe's specimens of the first fasciculus, taking them in their order, are these :— * No. 88. S. cinerea, L., var. B, Koch. S. aquatica, Sm.” From Yorkshire. Andersson thought that, as regards the leaves. at least, this tended towards S. Jaurina. It is rather a puzzling plant and suggests aurita x Caprea. “No. 39. S. cinerea, Sm.?” From Yorkshire. Andersson thought that this was near S. Seringeana, Qaud. (=tncana x ei- nerea). So far as the specimens go it seems to me S. cinerea only. * No. 40. S. cinerea, L., B, Koch. S.. aquatica, Sm.?” From Essex. This seems to be cinerea x aurita. “ No. 41. S. aquatica, Sm.” From Essex. A form of cinerea, says Andersson. This also seems to be cinerea x aurita. * No. 42. S. aquatica, Sm. ?" From Essex. Andersson calls it a subspecies of S. cinerea. To me it seems certainly a good form of cinerea x aurita. , * No. 44. S. oleifolia, Sm. ? ” From Essex.: ‘A form of cinerea, according to Andersson. I should call it cinerea x aurita. In Fasciculus iv. * No. 84. S. aquatica, Sm.," seems to be 5. cinerea, and “ No. 108. S. cinerea, L., S. aquatica, Sm.," cinerea X aurita, In addition to these specimens I have examined a number of others which have been distributed by various botanists and labelled S. cinerea, S. aquatica, and S. oleifolia, as the case may be. These, just as the above, tend to show that there exists 12 the minds of botanists a considerable vagueness as to how the Smithian names should be applied. As, therefore, there is now no certainty about them, it will be expedient to drop the names aquatica, Sm., and oleifolia, Sm. 6. SALIX AURITA, L. Contrary to what is the case in the majority of British gpecies REVISION OF THE BRITISH WILLOWS. 383 of Salix, the continental salicologists have described more forms of S. aurita than British botanists have done. |. Thus Andersson specifies three, and Wimmer four, principal modifications ; but as these are so connected by other intermediate forms that it is often impossible to say to which a specimen should be referred, it seems unnecessary to mention them by name. ` S. aurita, like its allies, is subject to a great range of variation. At the same time it can usually be recognized without much difficulty, not so much by any one characteristic as by a combi- nation of them and its general appearance. Its closest affinity is with S. cinerea, and, so far as dried leaf-specimens go, it is sometimes not easy to determine to which species they belong, since certain features of the twigs and buds, described as charac- teristic of the species, will, in practice, be found not quite reliable. Thus the year-old twigs, which ought to be glabrous, are not unfrequently more or less slightly pubescent in S. aurita, and glabrous when they ought to be tomentose in S. cinerea and the buds are glabrous or pubescent, though described variously as one or the other. From this close relationship it follows that the hybrids between S. aurita and S. cinerea, when not exactly intermediate but more related to one or other of the parents, can only be distinguished with great difficulty. X SALIX LUTEsCENS, 4. Kern. (S. cinerea X S. aurita.) Wimmer remarks that the hybrid forms—which he calls S. qurita-cinerea— between S. cinerea and S. aurita are most difficult to make out, since the differences between the species themselves are not easily expressed in written notes, but that, nevertheless, such hybrids seem to be commoner than has been thought. If the student, however, has familiarized himself with the two *pecies in question, so as to understand, in some degree at least, the range of their variation, the diffieulty of recognizing the hybrid forms is not insurmountable when they occupy a more or less intermediate position. When, however, the hybrid shows, as is very frequently the case in such plants, greater affinity with one parent than the other, absolute certainty becomes well-nigh Impossible, SCH Points of distinction between S. cinerea and S. aurita to be Chi iefly keptin mind are as follows :—twigs, in aurita more slender 2»2 884 DR. F. BUCHANAN WHITE'S and glabrous, in cinerea stouter and pubescent; leaves, in aurita smaller, softer, and more rugose; catkins, in aurita much smaller and with more distinct leafy bracts at the base ; scales, in aurita narrower, more ferruginous, and less black at the tips ; capsules, in aurita smaller, whiter, more pubescent, less subulate and more cylindrical, with no style and with short stigmas. Theoretically the hybrid should have characters altogether intermediate, but practically it will be found that in some points the cinerea-influence predominates, and in others the aurita. The twigs are usually more slender than in cinerea and either pubes- cent or glabrous: the leaves are very variable, and from them alone it is impossible to determine the hybrid; but compared with those of cinerea, they are usually smaller and show, especially in a living state, more rugosity, whilst their shape and general appearance suggest a mixture with aurita; the catkins are inter- mediate in size and shape ; the scales, whilst retaining a resem- blance to those of aurita, are blacker at the tips; the capsules show an evident relationship with aurita in their whiter colour and more cylindrical form, whilst, in being more evidently though very shortly styled, they betray affinity with cinerea. In the d plant reliance can be placed only in the intermediate size of the catkins and nature of the scales, taken, of course, in combination with twig- and leaf-characters. Whilst these points indicate generally what is to be expected in the hybrid, every specimen must be judged on its own merits, and due weight allowed to certain features indescribable, but yet easily recognized by the practiced eye. In Britain I believe that Salix lutescens is not an uncommon plant. Though British authors seem to have overlooked the record, Wimmer, so far back as 1866, mentions that he had seen specimens from Coventry Park, Warwickshire, collected by T. Kirk and distributed as S. cinerea var. oleifolia, Sm. This record of Wimmer's is suggestive of what turns out to be really the case, that in British herbaria S. lutescens has often been named S. cinerea or S. oleifolia, or even S. aquatica. As has been mentioned under S. cinerea, several of Leefe's published specimens of this group must be referred to S. lutescens, Some of these were issued as S. aquatica or S. oleifolia; but I think that in Fasc. i. Nos. 40, 42, and 44 (all from Essex), and in Fasc. 1Y- No. 108 (from Northumberland), must be called more or less good lutescens, as is possibly also No, 41 (of Fasc. i). As noted REVISION OF THE BRITISH WILLOWS. 385 also under Salix cinerea, Smith's own specimens of S. oleifolia are possibly hybrids of aurita and cinerea, but will have to be compared with authentic examples of lutescens. In Perthshire, whilst, in several places, specimens of whose hybrid origin there is no doubt, and which show several gra- dations, occur, I have seen a number of others about which there is much uncertainty. In places where both S. cinerea and S. aurita grow, a number of intermediate forms can be ob- tained, but plants of what also appears to be S. lutescens occur along with S. cinerea only. These, however, grow on the banks of rivers, and it is probable that they have been brought thither by water. Along with them grow other plants which differ very slightly from true cinerea and which may be (though this requires proof) hybrids of cinerea with lutescens. In the meantime I prefer to consider them as untypical cinerea. T have also seen specimens of S. lutescens from other parts of Britain, as, e. g., from Primside Bog, Roxburgh (A. Brotherston), Thirsk, Yorkshire (W. L. Notcutt), Crabtree, Devon (Archer Briggs)—alllabelled as cinerea ; named oleifolia, Sm.—Thirsk, Yorkshire (J. G. Baker), lutescens on the cinerea side; Quintin Pool, Warwick (T. Kirk), very near aurita ; named aquatica, 8m. —Falkenham, Norfolk (W. L. Noteutt); Hatton, Warwick (2. Bromwich) ; and Dorset (Salter). Other specimens, which I refer to lutescens, are from Caithness (E. F. Linton), Worcestershire (R. F. Towndrow), Kincardineshire (Trail), Clova, Forfarshire, and Derbyshire (W. R. Linton), and Surrey (W. H. Beeby). tis probable, therefore, that S. lutescens is a species widely distributed in Britain. At the same time it may be somewhat local, since S. cinerea and S. aurita, though not unfrequently associated, seem to prefer habitats of a rather different nature— Wet moorlands in the case of aurita and river-banks in that of cinerea. Wherever the two do grow in proximity (and such places occur in many districts) there S. lutescens should be looked for. 7. SALIX Caprea, L. sra e " Monographia ' Andersson described a number of forma nor doc. rs most of which are not mentioned in the ‘ Prodromus, In Bae immer name any varieties. . ite ch rain the species is, for a willow, so tolerably constant in Characters that it is not likely to be mistaken for any other. 886 DR. F. BUCHANAN WHITE'S At the same time it has a certain range of variation. The year- old twigs and the buds, which are normally glabrous, are not unfrequently slightly, and sometimes decidedly, pubescent ; the leaves, while typically roundly-oval, vary both in size and shape, being sometimes oblong, and at others much attenuate at each end—this latter form being, perhaps, more frequently found in northern and mountainous districts; and the catkins, though usually sessile or subsessile, are not very rarely provided with a leafy peduncle. Of this latter form 1 have found some rather curious plants in Perthshire. These, in addition to having the rather smaller catkins (in both sexes) furnished with a conspicu- ously leafy. peduncle, form dwarfer and more slender bushes than is usually the case with Salix Caprea, and are later in flowering. The leaves are, however, not dissimilar to the ordinary form of the plant, and hence are not referable to the modification or variety S. sphacelata, Sm. S. sphacelata, Sm., is now considered to be a subalpine form of A. Caprea, though Wimmer, from the figure and description, thinks that it may be referable to S. silesiaca. As well as differences in the leaves, it is said to have smaller catkins than in S. Caprea; but, whilst for these reasons it seems to be a rather well-marked form, I have seen too few specimens to be able to come to an opinion about it. The examples in Smith’s herbarium (“from Mr. E. Forster's garden ") look to me somewhat like a hybrid between S. Caprea and S. aurita. So in some respects do speci- mens labelled * Salix sphacelata, Sm., Cult. Hort. Kew. J. G. Baker ;” but these agree still better with Wimmer’s specimens Coll. No. 185) of S. Caprea-cinerea (=S. Reichardti) and, at any rate, are not pure S. Caprea. X SALIX REICHARDTI, A. Kern. (S. Caprea x S. cinerea.) S. Reichardti, A. Kern. (S. Caprea-cinerea, Wimm.), is & hybrid between S. Caprea and S. cinerea, which, according to Wimmer, is—like other hybrids of this group—very difficult to recognize on account cf the close affinity of its parents. As in Britain both S. Caprea and S. cinerea are common species, the hybrid between them might be expected to occur not rarely ; but such does not seem to be the case, perhaps because their habitats are not quite identical (Caprea being a chiefly woodland, REVISION OF THE BRITISH WILLOWS. 387 and cinerea a river-bank species), but more especially because their periods of flowering are not exactly synchronous. At the same time I think that the hybrid does occur, since T have seen a few specimens which, if not to it, I know not where they belong. In Perthshire I have found several plants which, though not good intermediate forms (some leaning to Salix Caprea and others to S. cinerea), I can refer only to S. Reichardti. Other specimens are from Fifeshire (in Edinburgh University Her- barium), Worcestershire (R. F. Towndrow), and Kent (E. S. Marshall). A plant from Kew Gardens, labelled S. sphacelata, Sm., seems also to belong here; and a leaf-specimen from the « towing-path near Kew, Surrey” (E. De Crespigny) has leaves which are quite intermediate between S. Caprea and S. cinerea. As Wimmer indicates, it is very difficult to point out in words the characteristics of this hybrid; and unless the student knows well the essential features of Caprea and cinerea, he will scarcely succeed in recognizing it, but rather place it as belonging to one or other of its parents, between which, of course, the combinations may be various. On account of the plate in Sal. Wob. (t. 127) and Dalle opinion regarding Eng. Bot. t. 1437, Wimmer thinks that S. aquatica, Sm., is a synonym of his S. Caprea-cinerea ; but, a8 pointed out under S. cinerea, I am inclined to believe that aquatica isa hybrid of aurita with probably S. Caprea. X SALIX CAPREOLA, J. Kern. (S. Capreax 8. aurita.) What has been said regarding the difficulty of identifying the hybrids of S. Caprea with S. cinerea, applies with nearly equal force to the hybrids of S. Caprea and S. aurita. Like the former they, too, are not of common occurrence, since the periods of owering of the species are not quite identical. In a locality near Perth, where S. Caprea, S. aurita, and S. cinerea, as well as the hybrid between the two latter, occur to- gether, I have found sume plants which I believe are S. capreola, though of different forms and not quite like any of Wimmer's published specimens. One bas slender branches resembling those of 8. aurita; leaves thin, intermediate between S. Caprea and S. rurita, and not unlike the Sal. Wob. figure of S. aquatica ; cat- ms (9) large ; capsules subulate from a broad base, with a short 388 DR. F. BUCHANAN WHITES style and very short erect stigmas. Though apparently a hybrid of Salix Caprea with S. aurita, it is just possible that it may be with S. lutescens (cinerea X aurita)—which is common in the same locality—from certain points of resemblance which it has to 8. cinerea, Another plant from the same place is a rather different form. The slender aurita-like twigs are more pubescent (in some of Wimmer’s examples they are pubescent); leaves more obovate in shape, near aurita, but with some trace of S. Caprea in them ; catkins ( 9 ) near aurita, but with black-tipped scales; capsules distinctly styled, stigmas longer and not so constantly erect. In Worcestershire Mr. R. F. Towndrow has found, along with less well-marked forms, one bush which is a very pretty condition of capreola, and notable for the distinctness with which it exhibits its hybrid origin. The catkins (9) are in facies like those of aurita, but the ovaries are more subuiate, the style more evident though very short, and the stigmas longer; the twigs are those of aurita, but the leaves are nearly intermediate, leaning a little to the Caprea side. Still another form is presented by plants collected at Clevedon, N. Somerset, by Mr. J. W. White and labelled S. aquatica. Of the parentage of these there seems to be little doubt. The connection with aurita is shown by the slender twigs, the shape in some degree of the leaves, and very short stigmas ; while from Caprea has been derived the larger catkins ( 9 ), the short but distinet style, the pubescence, veining, and, to some extent, the shape of theleaves. In the d plant the only catkin which I have seen (of these Somerset specimens) is nearer that of aurita. A willow ( 9) from Trysall has narrower leaves than those just mentioned, but seems to be another form of capreola. In addition to these I have seen more or less satis- factory examples from Derbyshire (W. R. Linton), Surrey (W. H. Beeby), and Kent (E. S. Marshall). A puzzling plant from Hurstpierpoint, Sussex (F. A. Hanbury), must also be referred to this hybrid. It has slender glabrous twigs, intermediate in character; leaves (rather young) also intermediate, but not distinctly recalling either aurita or Caprea ; catkins (rather old) larger than aurita; scales of aurita; capsules small and shortly styled. Still another form occurs on rocks on the west bank of the river Naver, Sutherland (F. J. Hanbury), which, though most probably Salix capreola, must, as it has no flowers, remain at present a little doubtful. The facies of the plant is that of 3 e er SEP ree REVISION OF THE BRITISH WILLOWS. 889 small form of Salix Caprea ; but the leaves are obovate, conspicu- ously stipuled, rather rugose, and not so pubescent as in Caprea. Much resembling this last are specimens, also flowerless, from Glen Dole, Clova (B. Daydon Jackson). Their leaves are much like some of Wimmer's specimens, but in the absence of catkins it is impossible to say whether the plant is S. capreola or a variety of S. Caprea. The probable identity of Smith's specimeus of S. aquatica with aurita x Caprea has been already mentioned under S. cinerea, and I rather suspect that Leefe's No. 38 (Fasc. i. “ S. cinerea var. p, Koch ; S. aquatica, Sm."), from Yorkshire, is possibly another form. Andersson thought that in the leaves at least the relation- ship of this specimen was with S. laurina (= Caprea x phylicifolia), but from this opinion Ward (who collected the plant) strongly dissented. The plant is a very puzzling one, but, from the shape and veining of the leaves and the pubescence of the younger ones (the older being subglabrous), it may possibly be Caprea x aurita. The catkins ( ¢) are large and not much is to learned from them. In continental Europe S. capreola, like S. Reichardti, is not a Common species. Wimmer, who describes five forms, attributes to it a wider range of variation than does Andersson, and from Its hybrid origin this it should naturally present. Like other hybrids, its characters consist in a combination of those of its parents; and as these are more or less variable, so also will be the resulting combination. Each specimen, therefore, which is Supposed to belong to S. capreola must be judged op its own merits, Group 5. REPENTES. 8. SALIX REPENS, L. ` repens is one of the most variable, in all its parts, of Euro- ge "Dun, and hence several species were made out of it by duced t, salicologists. These supposed species were soon re- retained the rank of varieties; but even as such they cannot be is, of « » Since none of their characters are to be relied on. It escri Weis possible to find specimens which agree with the deel lons; but, on the other hand, many examples combine in satisf ves the characteristics of several forms, and cannot be acto rily referred to one more than to another. * Principal modifications Wimmer gives a. argentea, b. fusca, 390 DR. F. BUCHANAN WHITE'S c. vulgaris, and d. rosmarinifolia, Koch (non L.). Andersson, who started in the * Monographia’ with three chief forms (repens, fusca, and arenaria) and one subspecies (Salix rosmarinifolia, L.), maintains in the * Prodromus" the latter only, and that as a variety, distinguished by its globose catkins and much longer linear leaves. It is doubtful, however, whether it deserves special mention, though in its extreme condition it is well marked. But be that as it may, the S. rosmarinifolia of British lists requires somè notice. This made its first appearance in British books in Hud- son's ‘ Flora Anglica, 1762. Hudson quotes asa synonym “ Salix pumila Rhamni secundi Clusii folio, R. Syn. 447,” in citing which he is followed by Smith (* English Flora’), who, however, gives as the authority for that description “ Dill. in Raii Syn. 447,” and says of the plant, “ Found by J. Sherard. Dill.” Though Hudson gives no locality for his plant beyond “Habitat in montosis udis," it seems probable that both Hudson's and Smith's records are founded upon Sherard’s specimen in the Dillenian herbarium; but Smith says that it was also “sent by Mr. Dickson, probably from Scotland, to Mr. Crowe." Since Smith's time S. rosmarinifolia has continued to appear in our books, as, for example, in Hooker’s ‘British Flora,’ ed. 4, 1838, and Hooker and Arnott’s ‘ British Flora,’ ed. 8, 1860—in both of which Sherard and Dickson are recorded as the only finders ; Babington’s ‘ Manual,’ ed. 8, 1881, where "S. P" is given as the distribution ; and Hooker's «Student's Flora, ed. 3, 1884, in which it is placed as a form of S. repens, “said to have been found in the last century by Sherard in bogs in Scotland,” ** Sherard ” being evidently a slip of the pen for * Dickson." In addition to the descriptions the plant has been figured in Eng. Bot. t. 1865 and Sal. Wob. t. 87 ; and Eng. Bot. t. 1366 and Sal. Wob. t. 86 have been supposed to represent avariety of it. Moreover, specimens have been published by Levfe, namely Sal. Exs. i. No. 19, * Salix rosmarinifolia, L., E. B. t. 1365," * Received from Mr. Borrer many years ago, but not as a British species," and No. 24,“ Salix rosmarinifolia, L., S. Arbus- cula, Forbes, Sal. Wob. t. 86." Through the kindness of Mr. G. C. Druce, I have been able to examine the willows of the Dillenian Herbarium. In it the “ Salix pumila Rhamni secundi Clusii folio,’ Ray Syn. 447. n 2, “found amongst Mr. J. Sherard’s plants, the place not named,” is a very bad specimen with old 9 catkins and young leaves REVISION OF THE BRITISH WILLOWS. 391 Though the leaves are rather narrower and more crowded than is usually the case, there seems to be no doubt but that the specimen belongs to Salix viminalis. Mr. Druce thinks that another willow in the same herbarium-—the “ Salix pumila angustifolia inferne lanuginosa” (J. Bauhin) Ray Syn. 447. 2—was also referred by Smith to S. rosmarinifolia. In this case the plant is S. repens. As to the S. rosmarinifolia of Linné, salicologists differ in opinion. Andersson considered that it was the above-mentioned narrow-leaved and globose-catkined form of S. repens, and I am inclined to agree with him. Wimmer, on the other hand, believed it to be the hybrid between S. repens and S. viminalis (S. vimi- nalis-repens, Lasch, S. Friesiana, And.) Wimmer referred Æ. B. t. 1365 (S. rosmarinifolia, * Linn.," Koch, Boswell-Syme) and Sal. Wob. t. 87 (S. rosmarinifolia, Sm.) to S. viminalis-repens, and by the evidence afforded by Leefe's specimens (Sal. Ess. i. No. 19) mentioned above, Wimmer's identification seems to be correct. The other plates, Eng. Bot. t. 1366 and Sal. Wob. t. 86 (S. Arbuscula, Sm. not L.),appear to represent the form rosmarini- folia of S. repens, as do Leefe's specimens No. 24, for which he quotes Sal. Wob. t. 86. Boswell-Syme called Eng. Bot. t. 1366 “S. rosmarinifolia var. angustifolia, Wulf.;" but Wulfen's S. an- gustifolia is merely a synonym of the rosmarinifolia form of repens. In Smith's herbarium the specimen labelled “ S. Arbuscula, Fl. Brit., Mr. Crowe's Garden, 1804,” seems to be the same thing. To which of the forms Dickson's plant *sent to Mr. Crowe" belonged it is now impossible to say; but I suspect that from what Forbes, W. J. Hooker, Borrer, and Walker-Arnott say regarding the close affinity of rosmarinifolia, Sm., and Arbuscula, Sm., it was probably a repens form. 1 think that in the meantime neither S. repens f. rosmarinifolia nor A viminalis-repens can be admitted to a place in the British Flora. In Linné’s herbarium an example, labelled “ 22 incubacea ” by Linné himself, and, in an unknown hand, “ Salix latifolia minor, C. B. 474," is a slender straight-branched repens with oblong straight-pointed leaves moderatel y pubescent below. Some authors have referred S. incubacea, L., to S. repens x aurita, and others to S. repens x viminalis. _ It remains to be noticed that the specimens of Leefe's Sal. Exs. 1. No. 11, « g, fusca, L., var. parvifolia, E. Bot. t. 1961 ? ; Salict. Wob. 81,” with the remark “This is remarkably devoid of the 392 DR. F. BUCHANAN WHITE'S silky pubescence characteristic of its tribe*. Received from Mr. Darwell, to whom it was sent by the late Mr. Forbes," certainly belong to Salix Arbuscula, L. Sal. Wob. t. 81 is, as regards the d plant, in agreement with Leefe's specimens ; but the 9 plant figured is doubtful, and is more like S. repens. Sal. Wob. t. 80 (S. adscendens, Sm.) is also probably S. Arbuscula, L. The Eng. Bot. figures, however, represent S. repens. Under S. repens must be placed, in the meantime at least, à willow in Mr. F. J. Hanbury's “ Boswell herbarium "' labelled “ Salix fusca, Mullaghmore, Co. Sligo, J. B. Syme, 1840.” The specimens are old and not in first-rate condition. The leaves are elliptic in shape, and hence not dissimilar to some forms of repens, but of a rather paler green and, almost from the first, glabrous both above and below, though some of the young leaves are densely silky and all are more or less ciliate on the entire margin. The catkins ( 9 ) have leafy peduncles (whose leaves have buds 10 their axils) from 4 to 1 inch in length; the capsules are moderately hairy, pedicellate, acute, with (compared with ordinary repens) remarkably long and slender styles. The specimens much resemble Wimmer's specimens (Coll. No. 254) of S. repens-myrtilloides, Wimm. (S. Aurora, Lestad., S. finmarchica, Willd.), though not identical with them. Koch makes S. finmarchica, Willd., a variety (e. finmarchica) of S. repens, with glabrous leaves, the young ones silky, and glabrous capsules. It is much to be desired that this plant could be rediscovered, as it deserves further study. X SALIX AMBIGUA, Ehrh. (S. repens x S. aurita.) S. ambigua, beyond doubt a hybrid of S. repens with S. aurita, is, according to Wimmer, the most frequent of all the hybrid willows. In Eng. Bot. (8rd ed.) four varieties are enumerated, viz.:— a. genuina, B. major, y. spathulata (Willd.), and à. undulata. Wimmer mentions two forms (in addition to the more typical state), B. microphylla and y. longifolia; and Andersson gives two modifications, 1. S. spathulata, Willd. (more related to S. aurita), and 2. S. plicata, Fr. (nearer S. repens). By Wimmer S. spathu- lata, Willd., is considered to be a synonym of S. aurita. * Although, perhaps, most usually pubescent, S. repens is not invariably 8°- I have found plants quite destitute of pubescence, but which, under cultivation, became more or less hairy. REVISION OF THE BRITISH WILLOWS. 398 Just as in other hybrids a number of specimens will not fit the varieties defined by botanists, but present intermediate cha- racters, so in Salix ambigua there is a continuous series of forms stretching from S. aurita to S. repens, and showing various degrees of combination of the characters of the two species. Hence, as it is impossible to say of all examples to which variety they should be placed, the expediency of retaining any varietal names seems doubtful. In Britain the distribution of S. ambigua is not well worked out, nor does the plant seem to be thoroughly understood. Though possibly not abundant, it probably occurs wherever $. aurita and S. repens grow at all commonly together, and as these species have been found almost throughout Britain, are very frequently associated, and flower at the same time, it is likely that S. ambigua has a wide distribution. It can often be re- cognized by the leaves alone, but most easily when these are intermediate. In less intermediate forms the greater silkiness of the pubescence serves to distinguish it from S. aurita, and the greater rugosity of the surfaces and less silky pubescence from S. repens. X SALIX CINEREA-REPENS, Wimm. (S. repens x S. cinerea.) l A willow found by Mr. F. J. Hanbury on sea-cliffs near Mel- vich, in Sutherlandshire, J uly 1886, has such a strong resemblance to S. repens that it might well be passed over as a curious variety of that species. The pubescence of the leaves is, however, different, being less silky and more crisped ; the leaves are more uniformly larger, as, perhaps, are also the catkins, and thus it appears to be a hybrid of S. repens with one of the Capree. The Smoothness of the leaves indicates that S. aurita is not one of the parents, and their resemblance to one specimen in Wimmer's examples (Coll. No. 245) of S. cinerea-repens suggests the probable Parentage. These Melvich specimens are, however, nearer S. “pens than most of Wimmer's. If more adult specimens as regards the leaves, and less mature as regards the catkins (9) poul be obtained, they might very probably show greater dif- erences from S. repens. At the same time the plant appears to ?6 with little doubt, S. cinerea-repens, Wimm. specimen (in Mr. A. Bennett’s herbarium) from Holme Fen, SEN (where cinerea and repens are the “most frequent wil- V8), probably also belongs here, though nearer repens. Though 394. DR. F. BUCHANAN WHITE’S I have little doubt regarding it, yet the specimen is too poor for absolute certainty, and is mentioned merely to call attention to the probable occurrence of the hybrid. A leaf-specimen of a willow from Clova (W. R. Linton) is also perhaps an example of this hybrid. X SALIX CaPREA-REPENS, Lasch. (S. repens x S. Caprea.) Very closely related to the last (and with it to S. ambigua, of which, indeed, Andersson is inclined to think they are both forms) is the hybrid between S. repens and S. Caprea. To it I think belong two plants found by the Messrs. Linton on the cliffs at Armadale, Sutherlandshire. Unfortunately there are no flowers on either of the specimens, and till these are seen there must remain a little doubt*. One of these specimens agrees with authentic specimens of S. Caprea-repens from Sweden, and is a fairly intermediate form. The other, whilst evidently belonging to the same hybrid, is, in the greater silkiness of the pubescence, rather nearer repens. Whilst these are undoubted hybrids of repens, the general facies and veining, pubescence, and colour of the leaves points to Caprea rather than cinerea or aurita being one of the parents, and therefore I give it a place in the list. X SALIX NIGRICANS-REPENS, Heidenr. (S. repens XS. nigricans.) Wimmer describes this hybrid from specimens sent to him by Heidenreich (who found two ¢ bushes near Tilsit, in Prussia), and compares it with S. ambigua. From that species it may be dis- tinguished by the dull green of the leaves as contrasted with the grey colour of those of S. ambigua, which are moreover wrinkled. Some willows found on the banks of the Tay above Dunkeld by Mr. C. M'Tntosh, and on the banks of the Garry, near Blair Athole, by myself, though not very like the only specimen of Heidenreich’s which I have seen, evidently belong to a hybrid between S. repens and S. nigricans. The specimens before me include both d and 9, and show much the same characters. In all respects they are intermediate * Since this was written I have seen a Ọ catkin grown by the Rev. W. F. Linton on a cutting of the better-marked of the two plants. This, while it has a considerable proportion of S. repens in it, has also part of its characters from S. Caprea, and tends to confirm the opinion I had already formed. REVISION OF THE BRITISH WILLOWS. 395 between their parents, though the leaves are suggestive of Salix nigricans rather than of S. repens. The d catkins are stouter than in repens, but smaller than in nigricans, the filaments slightly hairy at the base as in the latter, and the anthers with a tendency to become fuscous after the shedding of the pollen, as in the former. The 9 catkins are rather slender and, like the d , furnished with leafy peduncles; the remarkably slender young ovaries are glabrous or slightly silky towards the top, which passes gradually into the rather short style; the stigmas are short and resemble those of repens. The leaves in shape are not unlike some of the forms of repens, and are uniformly small, oval in outline, closely and finely crenate-serrate on the margin, the short tips straight or twisted, the upper surface with adpressed hairs, and the under- side—especially of the younger leaves—densely covered with a silvery silk-woolly pubescence. The branches are rather flexuous, but more divaricate than is usual in repens, at first pubescent, afterwards glabrous and shining, and reddish- or yellowish-brown 1n colour. It is possible that a hybrid between S. repens and S. phylici- Jolia (=8. Schraderiana, Willd., which is known only as a culti- vated plant) also occurs in Britain; but more specimens must be seer’ before it can be recorded. Group 6. PHYLICIFOLIS. The British plants of this group are S. phylicifolia, S. nigricans, s Arbuscula, and several hybrids of these with other willows. In diseussing the group, the first point to be considered is the Very difficult question of the rank as species of S. nigricans and S. phylicifolia, All authors are agreed as to the intimate alliance of these two forms, but no recent botanist, with the exception of Bentham, as ventured to unite them, though some have expressed their doubts regarding the specific distinctness. On the contrary, they have laboured to discover points of distinction which would, at all times, serve for the determination of these willows. Un- fortunately » however, for the student, these supposed distinctive Characters are not always assigned to the same species, and, 396 DR. F. BUCHANAN WHITE'S moreover, they are occasionally not in accordance with the authors' own diagnoses. In their extreme forms Saliz phylicifolia and S. nigricans can be separated without any difficulty, but a by no means incon- siderable number of examples exhibit such a combination of the characteristics of each, that it is impossible to determine under which name they should be placed. It may be the case that these perplexing forms are, as some authors suppose, hybrids between two distinct species; but, in view of the polymorphic nature of both S. phylicifolia and S. nigricans, it seems more probable that they are only intermediates which connect the extremes of one most variable willow. Adopting this view, I combine them under the name of S. phylicifolia, L., which seems to have originally included both. 9. SALIX PHYLICIFOLIA, Z. Whilst adopting the Linnean name for the series of European species which have been made out of S. phylicifolia, two subspecies or major varieties, namely a. S. phylicifolia, L., Auct., and f. 8. nigricans, Sm., should be distinguished. . The chief distinctions between phylicifolia and nigricans lie m the leaves andtwigs. In S. phylicifolia the leaves are thicker and of a firmer texture, of a brighter and more shining green on the upper surface and often more glaucous below ; whilst in nigricans they are thinner, less compact in substance, of a duller green and less shining above, and usually less brightly glaucous below—in both forms the underside may be green. S. nigricans has also A greater tendency to turn black in drying, but this is by no means invariable, and is of no real value as a characteristic, since some nigricans forms do not change colour and others appear to do so always. The leaves of S. phylicifolia are not only less pubescent (some times, indeed, perfectly glabrous), with the pubescence quite oT nearly disappearing at an earlier stage of their growth, but the hairs are usually —not always, perhaps—of a different character from those of nigricans. In S. phylicifolia the hairs are rather stouter, shorter, and straighter, somewhat shining, and, though mostly white in colour, bave a mixture of bright red-brown ones. In nigricans the hairs, which are more numerous and sometimes REVISION OF THE BRITISH WILLOWS. 397 very abundant, are softer, duller-coloured, longer, more slender and less rigid, and (always?) unicolorously whitish. As regards the twigs, the youngest—in both forms —are some- what pubescent, but more constantly and decidedly in Salix nigricans than in S. phylicifolia ; and whilst in the latter they are soon glabrous and shining, in the former they frequently remain downy and dull. In the organs of fructification there seem to be no sufficiently constant differences. But, as already mentioned, there are a considerable number of examples which show a greater or less combination of the charac- ters just described. In some the affinity with S. phylicifolia is exhibited by the more slender, brighter-coloured, and glabrous twigs; in others it is these parts only which retain the resemblance to S. nigricans. In intermediate forms the twigs are usually more phylicifolia-like, the leaves thick and firm, shining above, but less brightly green than in phylicifolia, somewhat, but not very, pubescent below, and with the shining and coloured hairs of phylicifolia mingled with the duller and softer pubescence of nigricans. Wimmer and others have considered these forms to be hybrids between phylicifolia and nigricans, and have distinguished them by the name of a phylicifolia-nigricans, Wimm. It may be that Wimmer's views are correct, and that these are really hybrid forms. I am inclined, however, to think that they are only the intermediate conditions of one species ; and if this is the case one name only—§, phylicifolia—must be retained, to the exclusion of 8. nigricans and $. phylicifolia-nigricans even as varieties. But Since the general concensus of opinion seems to be in favour of keeping nigricans as a form distinct from phylicifolia, it will be expedient, in the meantime, to distinguish by Wimmer's name, though not in his sense, the intermediate forms, and to use the designation S. phylicifolia-nigricans for those examples which show a combination of the characters of phylicifolia and nigricans, Having thus indicated the characteristics of the three forms Which I have united under S. phylicifolia, some of the British Varieties which were ascribed to each when they were treated as distinct species must be noticed. Of both a. Phylicifolia and S. nigricans a considerable number œt named forms. once supposed to be distinct species—are MINN. JOURN.—BOTANY, VOL. XXVII. 2E 398 DR. F. BUCHANAN WITTE!S retained in the British list. The characteristics of these consist in the shape of the leaves and in the amount of pubescence on the capsule and its stalk. An examination, however, of any consi- derable number of specimens will show that not only may the leaf-characters of one variety and the capsule-characters of another be present in one and the same specimen, but that both downy and glabrous capsules may occur in the same catkin. Leefe also has pointed out that under cultivation a smooth capsule may become pubescent. Whilst there is probably no doubt that the plants on which the supposed species or varieties were originally founded showed considerable distinctness, their characters are those of individuals, and these almost all cultivated specimens. Moreover, quite dif- ferent forms have not only been called by the same name by botanists who are supposed to have known the varieties, but have been figured and described. Since, then, these varieties do uot exist in nature as constantly distinct entities, the varietal names can be no longer retained. Of the varieties ascribed to Salix phylicifolia, one or two may be noticed more specially. S. tetrapla, Walker, being, according to Wimmer, the same à8 his S. phylicifolia-nigricans, has been considered by some authors to be, if not a hybrid, a connecting-link between phylicifolia and nigricans. But, as Wimmer himself points out, his specimens (received from the Berlin Botanic Garden) do not agree with the English figures. Leefe’s specimens (Sal... Ers. i. No. 8), “ re- ceived from Mr. Borrer as the plant of Walker," differ from Wimmer’s, and agree with Sal. Wob. t. 49, and appear, without doubt, to belong to a. 8. phylicifolia. Hence Wimmer's deter- mination (and the theories founded on it) of S. tetrapla, Walker, must be considered as erroneous, though his specimens seem probably referable to S. phylicifolia-nigricans. S. Croweana, Sm.—It is probable that under this name two plants are confounded—one an abnormal condition of phylici- folia, the other a hybrid of that species with S. purpurea. The essential characteristic is the combination of the.filaments of the stamens for a greater or less part of their length, as in A. rubra. But the union of stamens may originate in two ways--one DOF mally, by hybridization with the Salices Synandre; the other abnormally, by cladostemmy. In the latter the branches of the filament are said to form an obtuse angle, in the former an acute REVISION OF THE BRITISH WILLOWS. 399 one. From Smith's specimens, and from the drawings, the com- bined filaments of Saliz Croweana form acute angles, and hence are not strictly cladostemmic as that term is defined by Wimmer. From, however, the further abnormal development (as described and figured by Forbes, t. 52, and shown by specimens published by Leefe and Ward), when the stamens become changed into imperfect -ovaries, it seems probable that, as Borrer thought, the connate filaments of Smith’s Croweana were “ but an accidental monstrosity.” As such most of the specimens named Croweana must be considered, though some plants so called are phylicifolia without any deformity. Besides these, however, I have seen specimens of two plants which have been referred to Croweana, but which apparently derive their connate filaments from the hybridization of phylicifolia with purpurea. Since, however, these specimens are imperfect—having no mature leaves—and are, besides, not of certain British origin, it is unnecessary to say more about them on the present occasion. S. Dicksoniana.— According to Leefe, the original Dicksoniana, the plant described by Smith, “must be regarded as at present unknown.” efe refers his published specimens (Sal. Ess. i. No. 12) * received from Mr. Borrer as from Smith” to S. Dick- soniana, Forbes, t. 55. fig. 1, which they seem to be. Forbes him- self had doubts as to his plant being the same as Smith's, but apparently was not satisfied that it was distinct. Forbes’s figure has been sometimes referred to S. nigricans, to which it has some resemblance in the leaves; but Wimmer points out its likeness to Plants intermediate between S. nigricans and S. livida. Andersson cites Eng. Bot. t. 1390 under S. phylicifolia, but does not mention Sal. Wob. t, 55, except amongst the synonyms. All the speci- mens I have seen are cultivated ones, and probably from Borrer’s Plant. It is much to be desired that the form could be redis- Covered in a wild state. In the meantime, judging from the cul- tivated examples, I strongly suspect that S. Dicksoniana is à hybrid between S. phylicifolia and S. Arbuscula, since in its cha- Tacters it is almost intermediate between the two. , Wimmer does not distinguish by name any varieties of S. phyli- cifolia, but Andersson, in the * Prodromus, mentions two, which in the“ Monographia ' are treated as subspecies. These are S. Heget- sohweileri, Heer, supposed by Wimmer to be a hybrid of S. phyli- cifolia and S. hastata, and S. rhetica, Kern., which is described as a form of g, Phylicifolia approaching S. Arbuscula. 2x2 400 DR. F. BUCHANAN WHITE'S Tt remains to be noticed that Wimmer, adopting the opinion hat the Linnean Salix phylicifolia is doubtful, uses the name Weigeliana, Willd. The specimen in the Linnean Herbarium labelled by Linné “4, phylicifolia,” though perhaps nearer the phylicifolia of modern authors than itis to S. nigricans, is, on the whole, rather doubtful. We must now pass on to the varieties of 6. S. nigricans, Sm., which is one of the, if not the, most polymorphic of all Willows. On its modifications upwards of one hundred supposed species have been founded, eight of which are still retained, though as varieties only, in British lists. For the reasons already given, it is impossible to maintain these, though it is just possible that one or two of the mountain forms may yet be shown to have some claims to distinctness (as varieties) by their greater con- stancy of characters. Judging from the specimens published by Wimmer, Kerner, Billot, Reichenbach, &c., the Continental S. nigricans frequently differs less from S. phylicifolia than does the British plant. Many of these published examples have rather slender, glabrous or nearly glabrous, shining, chestnut-coloured twigs; and are distinguished from phylicifolia only by the thinner leaves, if even by that. It is not wonderful, therefore, that Andersson and others have spoken of the extreme difficulty with which the two “species " can be separated. Andersson, following Fries, gives two chief modifications of nigricans :—a. borealis (S. borealis, Fr.) and B. protea (S. cam- pestris, Fr.). S. borealis is a big bush or small tree, with stouter branches, more downy twigs, larger leaves, and larger catkins, with a leafy peduncle. „S. protea is a smaller bush, with thinner leaves and subsessile catkins. I have not seen authentic speci- mens; but, judging from the descriptions, both forms, as We, as intermediates, occur in Britain. In connection with this matter, that which has been said above with regard to the nature of the British nigricans as compared with the published Conti- Dental examples may be kept in mind. Wimmer mentions two varieties—(3. borealis, Fr., and Y. ma- crophylla, Hartig (with large glabrous leaves)—and notices a number of special forms, which, however, can scarcely be iden tified without named examples. y. S. phylicifolia-nigricans.—This name (afterwards altered to REVISION OF THE BRITISH WILLOWS. 401 Salix nigricans-Weigeliana) is that under which Wimmer describes the plant he received as S. fetrapla, Walker, but which, as already mentioned, appears not to be the true S. tetrapla. Wimmer plants, he thinks, were originally from Britain ; and the only habitat he gives for the supposed hybrid is Scotland. Under this name I place all the specimens which cannot be referred positively to phylicifolia or to nigricans. Such examples, though less common than those which can be referred without doubt, are not rarely to be met with, and in a few places are more abundant than either phylicifolia or nigricans. Andersson does not admit that there is any certain hybrid between S. phylicifolia and S. nigricans, and prefers to consider certain forms, which have been supposed to be hybrids, to be rather modifications of phylicifolia approaching nigricans, or of nigricans approaching phylicifolia. The Hybrids between the Phylicifolie and the Capres. Since in many parts of Britain species of the Phylicifolie and Capree groups frequently grow together, and their periods of flowering overlap, hybrid forms occasionally occur. It is not usually difficult to recognize these as hybrids, but it is not always easy to determine the exact parentage. Having reduced S. nigricans and S. phylicifolia to the rank of subspecies or major varieties of one willow, the number of the hybrid forms ought perhaps to be likewise reduced; but since the nigricans, or the phylicifolia, element in the compound is often distinetly marked, it seems expedient to keep their hybrids Separate. In their most intermediate forms the hybrids of this group show such a combination of the characters of their parents that their compound origin is at once evident; but, as usual, there are other forms which are not so easily recognized and are still more difficult to describe. The most certain character perhaps is in the structure of the style and stigmas—smalier than in the Phyli- cifolie, and larger than in the Capree; but there are also usually other good characters in the catkins and in the leaves. Whilst in most cases, though by no means in all, it is possible to decide whether it is nigricans or phylicifolia that is one of the Parents, there is often a very considerable difficulty in determin- Ing which of the Capree is the other, and more especially in those orms where the latter element is the less predominant one. 402 DR. F. BUCHANAN WHITE'S Consequently the proper position of a number of specimens must remain doubtful. That Salix nigricans hybridizes with Caprea and cinerea is ad- mitted by all salicologists; but hybrids with aurita have been denied by some, though it seems certain that they exist. Since, then, nigricans crosses with these three Capree, it seems but reasonable to suppose that phylicifolia should form analogous hybrids. Salicologists, however, allow only one such compound namely with Caprea. There occur in Britain, however, some willows which seem to show more or less distinctly the hybrid- ization of phylicifolia with cinerea and aurita. X SALIX LAURINA, Sm. (S. phylicifolia x S. Caprea.) XSaLix Warprana (Leefe, MS.) n. hybr. (S. phylicifolia X S. cinerea.) XSALIX LUDIFICANS, n.hybr. (S. phylicifolia x S. aurita.) X SALIX TEPHROCARPA, Wimm. (S. phylicifolia x S. cinerea X S. Caprea.) x Salix laurina. Regarding the position and limits of S. Jaurina, there has been much disparity of opinion. Whilst Borrer and W. J. Hooker thought that it was very distinct, Walker-Arnott was inclined to believe that it connects S. nigricans and S. phylicifolia, and that S. tenuior, Borr., and S. tenuifolia, Sm., were synonyms and varieties of it. Leefe thought that it could not be satisfactorily distinguished from S. phylicifolia, of which species Babington con- siders it as a variety. Boswell-Syme gives it as intermediate, and J. D. Hooker as a hybrid between 8. phylicifolia and S. Caprea. Andersson and Wimmer, on the other hand, have no doubt that it is a hybrid between pAylicifolia and Caprea ; though Wimmer says that, on account of the uniformity of the specimens which he had seen from various parts of Continental Europe, it is 4 question whether it should not be considered to be a distinct species,and nota hybrid. This constancy of form, he thinks, is due, however, to the fact that probably all the specimens have been cultivated from an English stock. Ihave seen too few European examples to venture to endorse Wimmer's opinion; but the British plants which have been referred to S. laurina are by no means uniform in their characters; and to this is probably due the differences of opinion amongst British botanists. This REVISION OF THE BRITISH WILLOWS. 403 absence of uniformity, moreover, arises from the fact that two, if not three, different hybrids have been confounded under the name of Salix laurina, Sm. Adopting Andersson's view that S. laurina, Sm., is a hybrid between phylicifolia and Caprea, the figure in Eng. Bot. t. 1806, the description in the third edition of that work, and the specimens published by Wimmer (Coll. no. 90) and Reichenbach (no. 2417) may be cited as illustrating the characters of the species*. In most respects it very greatly resembles S. phylicifolia (with which, Andersson says, it has often been confounded), but shows its relationship to S. Caprea, not only in the structure of the 9 catkins (the g is unknown), but in the subarborescent growth, and in the size (and pubescence when it is present) of the leaves. From the resemblance of the leaves, S. laxiflora, Borr., has been referred to S. Jaurina (as, e. g., in the * Student's Flora,’ and, with some doubt, by Andersson); and it is possible that it may be a state much nearer S. phylicifolia, though from the structure of the flowers it is more probably a form only of that species, x Salix Wardiana. In addition to plants which agree with S. Jaurina, Sm., as in- dieated above, others have been published which, while evidently closely related to them, seem to show affinity with S. cinerea rather than with S. Caprea. Amongst these are several published by Leefe, including that " received from Mr. Borrer many years ago as the plant of Smith” (Fase. ii. No. 38), as well as the following :— Sal. Brit. N o. 43, found near Richmond, Yorkshire, by Ward, and labelled “ T should refer it to aquatica. Borrer.” Of this An- dersson said “Mihi S. Jaurine forma;” and Leefeand Ward thought this a better verdict than Borrer's. The same plant was published 1n Sal. Exs.iii. No. 60, as S. phylicifolia, *S. lauring, Sm., proxima," With the remark that, though desiring to call it after the dis- ` à Eng. Bot, t. 1806 is cited by Smith under S. bicolor, Ehrh. (of which he 8 Mou, ern a synonym); but S. bicolor, Ehrh., is only 8. phylicifolia. rina » Forbes, t. 38, though it has Smith's description of his bicolor (lau- ) appended, seems to be a different plant; and Wimmer thinks that it “Presents phylicifolia, though the style is rather short. 404 DR. F. BUCHANAN WHITES coverer, Mr. Ward, he (Leefe) feared that it could not be satis- factorily distinguished from Salix lawrina. Specimens seem also to have been afterwards distributed with the MS. name * S. Wardiana.” Sal. Ess. i. “No. 3. S. laurina, Sm.” “sent me under the name of S. myricoides." Sal. Ess. iii. “No. 62. S. phylicifolia.” * S. laurine, Sm., proxima." * This connects No. 60 with S. Jaurina, Sm., No. 38, and S. myricoides, No. 3.” * That these specimens have a similarity to each other may be gathered from Leefe’s remark on the last (Sal. Exs. iii. No. 62), and that they also bear resemblance to S. cinerea, Borrer's note on No. 43 suggestively indicates ; though how that acute sali- cologist could refer the catkins to that species is rather puzzling. Leefe's specimens are all much nearer to S. phylicifolia than to S. cinerea ; but in Perthshire I have found several willows which complete the transition from one species to the other; and it 18 probable that a search in the localities whence Leefe’s specimens came would reveal similar intermediate forms. The Perthshire specimens show much greater resemblance to cinerea than to phylicifolia ; but that they are hybrids of cinerea with one of the Phylicifolie is demonstrated by the evident style, while the general glabrosity of the twigs, and to some extent of the leaves, indicates that phylicifolia rather than nigricans is the parent. Compared with the true S. laurina, this hybrid may be distin- guished by its smaller catkins and smaller leaves, the general appearance and shape of which are suggestive of cinerea. The pubescence of the underside of the leaves shows a combination of the characters of the hairs of phylicifolia and cinerea without the soft woolliness of that of Caprea. The capsules, which are somewhat variable, show no very great points of distinction, though usually smaller and without the yeilowish whiteness frequently exhibited by the Caprea hybrid. For the hybrid between S. phylicifolia and S. cinerea I have adopted Leefe's MS. name, S. Wardiana, since it commemorates a botanist who was a diligent and sagacious student of British * Besides these, Leefe published another, Sal. Brit. Ers." No. 73, S. lat- rina, Sm.,” found at Richmond, Yorkshire, by Ward. Of this, Andersson says that it recedes somewhat from the true 8. laurina, but is nevertheless near it. If it is laurina at all, it is very near phylicifolia, and I would rather refer the specimens I have seen to that species, REVISION OF THE BRITISH WILLOWS. 405 Willows. The name Salix Wardiana is, of course, used in a wider sense than that in which it was employed by Leefe. The d has not been noticed, and though it probably occurs, it is very possibly difficult to recognize. x Salix ludificans. In addition to the Willows just noticed, I have seen specimens from a few bushes, which, whilst undoubtedly very close to „S. phylicifolia, cannot be satisfactorily referred to that species, on account of their short styles and other peculiarities, which seem to indicate that they are of hybrid origin. The phylicifolia element in them is so predominating that it is difficult to deter- mine the other parent, but after considerable study I think it must be S. aurita. Since S. aurita flowers a little later than the other Caprea, it might be imagined that hybrids of it with the Phy- licifolie would be at least as frequent as those of the other Capree; but this is either not the case, or they have been overlooked. l For this hybrid I propose the name S. ludificans. As already indicated, all the specimens I have seen are very much nearer Phylicifolia than aurita, though, judging from other hybrids, it is probable that forms more remote from phylicifolia occur. The specimens and some of their distinguishing characteristics are as follows :— A bush found by Mr. C. M'Intosh on the banks of the Tay above Dunkeld has the catkins ( 9) quite intermediate between Phylicifolia and aurita; but the leaves are very near those of the former species, exhibiting, however, in their shape a sug- gestion of aurita. The scanty pubescence, moreover (present on some of the leaves only), shows an affinity with aurita, since the hairs are finer and softer than in phylicifolia and slightly crisped. The other specimens I have seen were collected in Caithness by Mr. J. Grant and sent to me by Mr. A. Bennett. The examples are, unfortunately, not in the best condition and the Material is scanty. They represent three bushes, of which one is Sand two?. One has the catkins ( 9 ) rather different from the above-mentioned Perthshire plant, but, like it, intermediate tween itg supposed parents, and has the young leaves (I have aot seen mature ones) apparently also intermediate. Another * a striking resemblance to some forms of the hybrid between 406 . DR. F. BUCHANAN WHITE'S Salix aurita and S. nigricans, but with the nigricans element re- placed by phylicifolia, and, though a different looking plant from either of the above, shows its hybrid nature in all its parts. The g, though, like the others, near phylicifolia, yet shows, in the small catkins and structure of the leaves, a relationship with aurita; but if it had not been for the 9 examples it would probably have been referred to phylicifolia. X Salix tephrocarpa. In connection with the hybrids of cinerea and phylicifolia, mention must be made of the Willow described by Wimmer under the name S. tephrocarpa. Of this one bush only seems to be known, and, though it grows in the Berlin Botanic Garden, its origin is quite uncertain. Both Wichura and Wimmer tried in vain to ascertain the parentage of this plant, and the latter finally came to the con- clusion that it might be a hybrid between Jaurina and cinerea. A Willow found by Mr. C. M‘Intosh on the banks of the Tay, above Dunkeld, agrees pretty well with the description of S. tephrocarpa ; but I have been able to compare the leaves only, which I owe to the kindness of Herr Hennings, Director of the Berlin Botanie Garden. The Berlin and Dunkeld leaf-examples, though not identical, have a fair resemblance to each other, and both show an undoubted relationship with cinerea. In other respects the Dunkeld plant bas some affinity with Caprea and, to a slight degree, with phylicifulia. The catkins (2) are large and very handsome, the white capsules contrasting strongly with the conspicuously black scales. The style is very short but present; and in this and some other points the affinity with phylicifolia is shown. In the meantime I can find no place for this form but as 4 hybrid between cinerea, Caprea, and phylicifolia, and possibly arising from cinerea x phylicifolia crossing with Caprea. X Butt LATIFOLIA, Forbes. (S. nigricans x S. Caprea.) Wimmer and, on his authority, Andersson consider that 5. latifolia, Forbes, t. 118, is a hybrid between S. nigricans and S. Caprea. Leefe, on the other hand, has published (Sal. Ezs. ii REVISION OF THE BRITISH WILLOWS. 407 Nos. 52 & 53), under that name and with citation of the plate, a plant which he says is only a form of Salix nigricans. Comparing Leefe’s specimens with Forbes’s plateand description, I find that they do not agree, and that whilst Leefe’s latifolia is nothing more than what he thought it to be, i. e. nigricans, Forbes's Plant is evidently a hybrid form. From the affinity between S. Caprea aud S. cinerea, their hybrids with S. nigricans frequently so much resemble each other that it 18 not easy to separate them. Wimmer relies on the yellowish- white much thicker capsules, thicker, shorter and broader catkins, the much broader, oval-subrotund young leaves, and the larger, broader, and more hairy old leaves, as characters by which to distinguish the best form of latifolia from hybrids of cinerea with nigricans, but remarks that some specimens show a departure from these points of distinction. S. latifolia seems to have been found in a very few places in Lapland, Sweden, and Germany. The only undoubted specimens I have seen are all * from Perthshire, where three bushes—in two widely separated localities—have been found by Mr. C. M'Intosh and myself. Of these, one is quite intermediate between S. Caprea and S. ni- gricans ; another greatly resembles Sal. Wob. t. 118 and has more affinity with the cinerea-nigricans hybrid, but also seems to be, beyond doubt, latifolia; and the third, in its longer styles, inclines More to nigricans. Of 8. latifolia the only was known to Forbes and to Wim- mer, but Andersson describes the d. ` Wimmer refers S. firma, Forbes, t. 106, and S. cotinifolia, Sm., F orbes, t. 114, with some doubt, to S. latifolia ; but Andersson thinks that they represent nigricans only. Of S. firma I have Seen specimens from Kew Gardens which, though not quite identical with the plant figured by Forbes, are probably the same ; they possibly represent another form of latifolia, but I think are rather cinerea x nigricans. The plant usually called in Britain cotinifolia does not appear to be the same as Forbes's, and 8 only nigricans. * e . . . Since this was written I have found in Edinb. Univ. Herbarium a specimen eege “S.W. corner of Duddingston Loch,” which must be referred to S. lati- 408 DR. F. BUCHANAN WHITES X SALIX srREPIDA (Schleich.), Forbes. (S. nigricans x S. cinerea.) Andersson uses the name S. puberula, Do, for the hybrid formed by S. cinerea with S. nigricans, because he and Wimmer think that S. vaudensis (Schleich.), Forbes, is, though probably the same thing, somewhat doubtful. Wimmer is also of opinion that S. strepida (Schleich.), Forbes, t. 100, bears more resemblance to the same hybrid than to anything else, though Andersson refers it to S. nigricans. Of S. strepida Y have received specimens from Kew Gardens which are sufficiently like Forbes's figure, and which seem, with- out doubt, a hybrid between cinerea and nigricans. Consequently, since the name is earlier than either vaudensis or puberula, I have adopted it for this hybrid. S. firma, Forbes, t. 106 (which Wimmer thinks may be S. Caprea-nigricans), and S. vaudensis, Forbes, t. 117, are, judging from specimens cultivated at Kew, forms of the same hybrid. Like other compound willows, S. strepida is subject to con- siderable variation, increased in this case by the variability of its parents. In addition to this, the intimate alliance, on the one hand, of nigricans to phylicifolia, and, on the other, of Caprea and aurita to cinerea, augments the difficulty of satisfactorily placing every specimen, and makes it impossible to draft such 4 description as will serve to identify the hybrid in every case- At the same time S. strepida, in many of its forms, has a facies of its own which, when once learned, should not fail to guide the student to a correct discrimination of the species. Wimmer and, following him, Andersson describe three forms— a. puberula (Doll), B. vaudensis (Forbes), and y. nitida, Wimm. From the examination of a number of specimens, I am unable to see any sufficient reason for maintaining these varietal names. Inits best forms, strepida combines the characters of its parents, but not unfrequently it exhibits more relationship with one than with the other. In some of its states the leaves so much resemble those of nigricans, that it is only by the shorter style and stigmas that any connection with cinerea can be suspected. In others the leaves, catkins, and capsules are so like those of cinerea that the relation with nigricans (or perchance with phylicifolia) is shown only by the rather long styles. In Continental Europe S. strepida has been recorded from ? REVISION OF THE BRITISH WILLOWs. 409 very few places only, perhaps because it has not been carefully looked for. In Britain it may not unreasonably be expected to occur in those parts of the north where its parents grow together, as they often do, since in Perthshire it is of wide, though not of abundant, occurrence on the banks of the Tay and some of its tributary streams. I have found both sexes, but have seen more ? than d plants, probably because the former are more easily recognized. In the south-east of Scotland Mr. A. Brotherston has found plants which probably belong to Salix strepida, though the material I have seen is scarcely sufficient for absolute certainty. X SALIX CORIACEA (Schleich.), Forbes. (S. nigricans x S. aurita.) In his notes on S. cinerea-nigricans (S. strepida), Wimmer says that it is hardly to be doubted that hybrids between S. nigricans and S. aurita occur. As examples of such he cites S. conformis, Schleich., in Herb. Willdenow, and thinks that Forbes’s figure, t. 119, of S. grisophylla (Schleich.), Forbes, also represents such a hybrid. Andersson, after quoting Wimmer’s words, says that the S. conformis referred to is represented by three specimens, of Which two belong to S. aurita and the third to S. cinerea, and that the specimens of S. grisophylla, published by Schleicher, Which he has seen, are, as regards some, S. nigricans, and, as regards others, S. cinerea. Under the name S. aurita-nigricans, Heidenr., Heidenreich has distributed specimens of a Willow found near Tilsit in russia; but I am not aware that any description of it has been published. Here and there, on the banks of the Tay between Dunkeld and Logierait, Mr. C. M‘Intosh and I have found plants which "ppear to be certainly a hybrid between S. nigricans and S. aurita. As might be expected, these bear a very considerable resemblance to 8. strepida, and some of them, indeed, might be nearly as well referred to that hybrid, except for a certain appearance suggestive H aurita rather than cinerea. Others, however, show a distinct combination of the characters of aurita and nigricans, and agree tolerably well with Heidenreich’s examples. Though Wimmer refers it to S. nigricans, | think that Forbes’s 410 DR. F. BUCHANAN WHITE’S t. 112, Salix coriacea, represents this hybrid. From the plate alone no certain conclusion could be derived; but specimens received from Kew Gardens, and others published by Leefe (under 8. nigricans), can scarcely be satisfactorily placed otherwise than here, though they do not exhibit the best form of the hybrid. In the plate both g and 9 flowers are shown, but I have seen the only. The flowers in the specimens show a more marked re- lationship with aurita than they do in the plate. If I am right in referring Forbes's plant, S. coriacea (Schleich.), Forbes, must be the name of the hybrid; otherwise it will have to be called S. aurita-nigricans, Heidenr. Like its close ally S. strepida, S. coriacea shows considerable variation. From that species the form of the leaves, recalling aurita rather than cinerea, the usually smaller and narrower 9 catkins, the shape and smaller size of the capsules, and frequently the colour of the scales are characteristic points which will serve to distinguish coriacea. So far, however, as our specimens go, it is more likely to be passed over as a form of nigricans, from which, however, careful examination will show its distinctness. A plant found on the banks of the Tay above Dunkeld by Mr. C. M'Intosh must be noticed here, since it seems to include nigricans and aurita amongst its parents. From certain points, however, in which it resembles the willow (from the same locality) already referred to as S. tephrocarpa, Wimm., I am inclined to suspect that it also includes S. Caprea and is a hybrid of that species with S. coriacea. 10. But AnBuscura, L. There is a divergence of opinion amongst botanists as to the situation of S. Arbuscula. Many, including the British, think that it should be associated with S. Myrsinites, &c.; but Anders- son, on account of the great similarity of the dwarfer forms of S. phylicifolia and the larger sub-alpine states of S. Arbuscula, considers that its place is amongst the Phylicifolia. À number of varieties or modifications have been described, but various authors have treated these differently. The varieties of the ‘London Catalogue’ are :—a. carinata, Su. b. fetida, Schleich. (=prunifolia, Sm.), c. venulosa, Sm., and Eta PNEU E * REVISION OF THE BRITISH WILLOWS. 411 d. vaccinifolia, Walker—all of which were at one time supposed to be distinct species. Wimmer has:—a. Waldsteiniana, Willd., 3. formosa, Willd., and y. feetida, Schleich. (to which he refers venulosa and vaccinifolia). Andersson has also three chief modifications—a. erecta (with three leaf-forms which include Waldsteiniana, formosa, and pru- nifolia); B. humilis (with two leaf-forms which include fetida, venulosa, and vaccinifolia) ; and y. thymeleoides, Schleich. (about which Wimmer is doubtful, and which Andersson thinks may be a subpilose condition). Both Wimmer and Andersson place Salix carinata, Sm., as a synonym of S. Arbuscula. Tf any varieties are to be retained, Andersson’s arrangement seems to be the best. S. erecta is distinguished by its more up- right and taller growth, its larger leaves, and less leafy catkins; humilis by its smaller size, more creeping habit, smaller leaves, and by the catkins which, when young, are subglobose and buried in leaves. But whilst plants occur which agree well with the definition of one or other of these modifications, there are many which can- not be placed in one more than in the other. I therefore think that it is inexpedient to adopt any varietal names. As for the British forms, there seems to be little doubt but that Walker- Arnott is right in saying that they cannot be satisfactorily dis- tinguished. There are some discrepancies in the various descriptions of S. Arbuscula. By some authors the style is described as elongate or long; by others as very short or mediocre, which in the majority of cases it is, The capsule is sometimes said to be sessile, whereas it has a pedicel of varying length but always much shorter than the rather elongate nectary, a character which distinguishes it from S. phylicifolia, whose pedicel is always longer than the short nectary. The stipules are described as absent or very rare; but small stipules may often be seen on the young “hoots. The coma of the seeds is sometimes described as having à reddish tinge ; but in all our specimens it is white. l Some other points in which S. Arbuscula varies may be briefly noticed. The leaves, which exhibit a considerable range in shape and size, vary also in the pubescence and colour. The underside -$ sometimes glaucous, sometimes green or with white dots ; Sometimes distinctly pubescent when young, and at others glabrous. young branches and buds are often silky, though frequently 412 DR. F. BUCHANAN WHITE’S described as glabrous. The scales vary in colour, pubescence, length, and shape (roundish, obovate, and oval) ; the style from nearly none to almost one third the length of the ovary, and, with the stigmas, in being thick or thin ; and the capsule in the amount of pubescence and colour. With all these variations there is, however, usually no difficulty in easily recognizing the species. But Dicxsonrana, Sm. (S. phylicifolia x S. Arbus- cula.) Although Wimmer was inclined to think that S. humilis, Willd., might be a hybrid between S. phylicifolia and S. Arbuscula, he does not, in the *Salices Europes,’ go the length of describing it as such, and Andersson says that it is a modification only of S. phylicifolia. Hybrids between phylicifolia aud Arbuscula might be expected to occur; but if they do they must be very rare. As mentioned under S. phylicifolia, it seems quite within the bounds of possi- bility that S. Dicksoniana, Sm., is such a hybrid. Though origi- nally found in “Scotland” by Dickson, and in Breadalbane by Winch, only cultivated specimens appear to be now known, and the characters of these, as well as those given in Smith’s description, are in many ways intermediate between phylicifolia and Arbuscula. As already stated (p. 399), there is some doubt whether Smith's plant and Forbes's are the same ; but the proba- bility seems to be that they are only different conditions of one species, and I have therefore retained Smith's name of S. Dick- soniana. As regards the specimens I have seen, the leaves might belong to either S. phylicifolia or to S. Arbuscula, though, perhaps, on the whole nearer the latter species. From S. Arbuscula, also, has been derived the small stature of the plant (*a foot high," Smith; “18 inches to 2 feet," Forbes), the small catkins (9) appearing with the leaves and with leafy peduncles, the small capsules, which in colour and aspect recall those of Arbuscula, and the shortish styles; but from phylicifolia the thicker catkins, more longly-pedicelled capsules with stronger pubescence, and stouter styles and stigmas. The scales are somewhat intermediate in colour and structure. Both Smith (who places it, in Engl. Fl., next S. carinata) REVISION OF THE BRITISH WILLOWS. 413 aud Forbes (next Salir prunifolia) seem to have had some thoughts of the reiationship of Dicksoniana to Arbuscula forms. Neither Smith's plate (in which the catkins are much too young) nor Forbes's (in which the figured leaves resemble those of nigricans) can be called very good. It is to be hoped that the wild state of Salix Dicksoniana will be rediscovered. Group 7. VIMINALES. 11. SALIX viMINALIS, L. . Though, like other species, liable to modification, S. viminalis Is, in Western Europe at least, one of the less variable willows, and has no varieties worth y of distinct names, though one or two have been described. Amongst these are var. intricata, Leefe (distinguished chiefly by the cloven, reflexed, and entangled stigmas), and var. stipularis, Leefe (with broader leaves and longer stipules). The 9 catkins vary a good deal in size. Most of the British and many of the foreign specimens I have seen have much shorter, but proportionately broader, catkins than in Wimmer’s published examples, but apparently do not belong to Déll’s var. abbreviata, in which the catkins are described as linear-oblong. XSatrx Surruraxa, Willd. (S. viminalis x the Capree.) The hybrids which S. viminalis makes with the Capree form à group the treatment of which is very difficult—a difficulty not diminisheg by the manner in which the group has been dealt with by botanists. ) For not only have different names been given to one and the “ame plant, but different plants have had the same name ascribed ‘ot em. Thus four, more or Jess distinct, forms have been called fcuminata,” and the true S. acuminata has been described under two other names. Much of this confusion has doubtless Teen, not Only from the great variability of the hybrids, but from the fact that the plants described by British authors were " familiar to the Continental salicologists, nor those of the atter to British botanists. No the 8roup is so specially a British one, it will be of interest later ne Its constituents according to the views of some of the T British botanists. It includes, according to :— LINN, JOURN.— BOTANY, VOL. XXVII. 2r 414 DR. F. BUCHANAN WHITE'S 1. W. J. Hooker and Borrer: S. stipularis, Sm., S. Smithiana, Willd., S. ferruginea, “ And., MS.,” S. acuminata, Sm., 8. holo- sericea, “ Willd.” ( 2S. acuminata, var. rugosa, Sm.). 2. Walker-Arnott: S. stipularis, Sm., S. Smithiana, WO, S. acuminata, Sm., S. ferruginea, And., S. holosericea, “Willd. ? 3. Leefe (Journ. Bot. 1871): S. stipularis, Sm., S. Smithiana, Willd., S. acuminata, Sm. 4. Boswell-Syme: S. stipularis, Sm., S. Smithiana, Willd. (with var. stipularis), S. ferruginea, And. (with var. rugosa), S. acum- nata, Sm. 5. Babington (1881) : S. stipularis, Sm., S. Smithiana, Willd. (with varr. rugosa and ferruginea), S. acuminata, Sm. 6. J. D. Hooker (1884): As supposed hybrids of S. viminalis— S. stipularis, Sm., S. Smithiana, Willd., S. acuminata, Sm., S. ferruginea, G. And., S. holosericea, Willd., S. rugosa, Leefe (7 holosericea, Hook. & Arn.). 7. London Catalogue (1886): S. stipularis, Sm., S. Smithiana, Willd. (with var. pseudo-stipularis, Lond. Cat.) S. ferruginea, G. And., S. rugosa, Leefe, S. acuminata, Sm. Wimmer has as the constituents :— S. Caprea-viminalis, Wimm. (=S. Smithiana, Willd.). S. cinerea-viminalis, Wimm. S. aurita-viminalis, Wimm. S. stipularis, Sm. S. Calodendron, Wimm. (—S. acuminata, Sm.). S. holosericea, Willd. Andersson praises Wimmer's arrangement, but remarks that authors have described individual specimens rather than the main forms themselves. His own arrangement is as follows. I have added within brackets the equivalents in Wimmer’s * Salices’ and in the * London Catalogue ' :— S. stipularis, Sm. (also of Wimmer and Lond. Cat.). S. Smithiana, Willd. «. sericans (Tausch). (S. Caprea-viminalis, Wimm. S. cinerea-viminalis, Wimm., p. P- S. aurita-viminalis, Wimm., p. P: S. Smithiana, Willd., Lond. Cat-) B. velutina (Schrad.). (S. cinerea-viminalis, Wimm., p- P S. holosericea, Willd., Wimm. S. rugosa, Leefe, at least p. p) REVISION OF THE BRITISH WILLOWS. 415 Var. ferruginea (Forbes). (S. cinerea-viminalis, Wimm. S. ferruginea, G. And., Lond. Cat.) y. acuminata (Sm.). (S. Calodendron, Wimm. S. dasyclados, Wimm. S. acuminata, Sm., Lond. Cat.) The weak point of this arrangement is that under one specific name (Sali Smithiana) several hybrid plants—having one only of the parents in common—are included. If S. Caprea, S. einerea, and 8. aurita are retained as distinct species—and Andersson did not unite them,—their several combinations with S. viminalis should necessarily, according to the method adopted by Andersson in the case of other hybrids, have separate recognition. On the other hand, comes the almost unsurmountable difficulty of distin- guishing between them, especially when, as in several cases, the exact parentage has not been proved, and when it is by no means improbable that a second hybridization, either with one or other of the parents or with a third species, has taken place. For this reason I am inclined to adopt provisionally the following modifi- cation of Andersson's arrangement :— S. Smithiana, Willd. a. stipularis (Sm.). B. sericans (Tausch). Y- velutina (Schrad.). d. Serruginea (G. And.). €. acuminata (Sm.). dat re proceeding to notice each of these varieties, I may say cimene ( e result of the study of a very large series of spe- tinental TH of them authentically named), both British and Con- will Wë , are failed to find such a permanence of characters as examples ° definitely separate one form from another. Certain under on can be, without much hesitation, placed satisfactorily ever that or other of the varietal names. There are many, how- anoth o cannot really be referred to one variety more than to of each , bi which, combining to some extent the characteristics said tha orm connecting-links. Of some others little more can be an that they are modifications of S. Smithiana. ie a. stipularis. (Salix stipularis, Sm.) d . ` pe the statement given above, it will be seen that most of the cited retain S. stipularis as a species. Even Andersson, Who . has united the other forms, gives it separate rank. But 2r2. 416 DR. F. BUCHANAN WHITE'S from a study, both of descriptions and specimens, I cannot find any grounds for giving Salix stipularis a higher position than that which some of the other forms (as, e. g., S. acuminata) have been considered entitled to. Of all the varieties, stipularis is the one which is most closely related to S. viminalis. From that species it may be distinguished by the broader leaves with less shiniug pubescence on the under surface, longer and broader stipules, larger catkins, and less sessile capsules; but whilst these are its normal characteristics, they may be all more or less so modified as to afford gradations into S. viminalis on the one side or into (9. sericans on the other. From the latter, its usually tomentose twigs, larger and darker catkins, more sessile capsules, and longer and more filiform stigmas serve to distinguish it, as do also, in the more typical examples, the shape and size of the leaves and of the stipules. But the latter are subject to modification and are not always present; and the former, which vary even in the same specimen, are not unfrequently so similar to those of some conditions of sericans that, were it not for the other characters, some examples would certainly be, and frequently are, called sericans. In a word, there is no sharp boundary between S. viminalis and stipularis nor between stipularis and sericans. Willows more or less resembling stipularis have been named S. viminalis var. stipularis, Leefe, and S. Smithiana var. pseudo- stipularis, Lond. Cat. Of these the first includes mere modifi- cations of S. viminalis, as well as forms of stipularis approaching viminalis; whilst “ pseudo-stipularis" seems to be applied to plants connecting stipularis with sericans. The d of stipularis seems now to be unknown * though it is figured both by Smith and by Forbes. The parentage, other than from S. viminalis, is doubtful, and some authors have thought that S. cinerea is the unknown factor in its origiD- . Wimmer suggests that it is a hybrid between S. viminalis and S. dasyclados, Wimm. (=S. acuminata, Sm.). I suspect that Wimmer is so far right in supposing that it has originated from a cross between S. viminalis and one of the viminalis-Capreé hybrids; but whether this hybrid is acuminata is doubtful. ` S. stipularis is almost confined to Britain, but, though 1 value as an osier is said to be small, whether it is really more than an escape from cultivation is very uncertain. * Sincethis was written I have found a single g bush—apparently self-sown— near Perth. I fear, however, that it has been lately eradicated. REVISTON OF THE BRITISH WILLOWS. 417 B. sericans. (Salix sericans, Tausch.) This, which includes “ Salix Smithiana as generally understood, is a most variable plant, and is almost inseparably connected with stipularis on the one side and velutina on the other. From the former, its smaller and paler-coloured catkins, longer- stalked capsules, and usually shorter and thicker stigmas, and from the latter its rather larger catkins and shorter-stalked capsules, are the best points of distinction ; whilst another, which separates it from both, is to be found in the usually glabrous twigs. But in any large series of specimens all these points will be found to be subject to much variation. The leaves are very variable, not only in shape and size, but in the nature and amount of the pubescence. In shape the leaves pass—even in the same example—from ovate-oblong to linear- lanceolate; whilst the pubescence of the underside may either be silky or woolly, shining or dull, copious or almost altogether absent. Asa rule, the margin is entire or very slightly crenate ; but in forms which approach velutina it is more evidently crenate or even subserrate. These latter forms Andersson places in modification 3. subobscura (of sericans). Some of the forms, as 1. latifolia, Aud., and 2. angustifolia (Wimm.), And., are, as Wichura has proved by experiment, hybrids of S. viminalis with S. Caprea; but the origin of others (as, e. g., 3. subobscura, And.) is not so clear. Under subobscura Andersson cites, as belonging in part, Wim- mer’s S. cinerea-viminalis and S. aurita-viminalis ; and from the characteristics of this form it seems very probable that S. cinerea and S. aurita are concerned in the production of some of its Modifications. Some of the specimens which have been referred to S. rugosa, Leefe, should probably find a position here. To these forms Andersson adds a fourth—serotina, which, however, he thinks is only accidental. It is distinguished by the catkins being shortly but distinctly stalked, and the peduncle furnished with small leaves. Many of our 9 plants have catkins of this nature. Var. sericans seems to be the most common of the varieties of 8. Smithiana, but appears to occur more frequently as a cul- tivated plant or as an escape from cultivation than as a truly wild Ze, The most abundant form of it is 2. angustifolia, or near it. It presents, however, great variability in small particulars. 418 DR. F. BUCHANAN WHITE'S y. velutina. (Salix velutina, Schrad., S. holosericea, Willd., S. cinerea-viminalis, Wimm. p. p., S. Micheliana, Forbes, Sal. Wob.) Var. velutina, which is often closely connected with sericans, and especially with the form subobscura (from some states of which it can scarcely be separated), is distinguished by its usually tomentose twigs, serrulate leaves, shorter style and stigmas, and more longly pedicelled capsule. Whilst in stipularis the capsule is scarcely stalked, and in sericans the stalk is about as long as the nectary, or at most twice as long, in velutina it is twice OT three times the length. Just as in the other varieties, however, all the characters are liable to modification; so that it becomes impossible to say where certain specimens should be placed. Though the leaves are normally more or less serrate, they are not invariably so. Forms of this group which have serrate leaves seem to have derived their origin from the hybridization of S. cinerea or S. aurita, rather than S. Caprea, with S. viminalis. This appears to be in a measure indicated by the position of the marginal glands. In S. viminalis the position of these is some what remote from the entire margin. In S. Caprea they are also rather remote from the margin, which, however, is by n9 means always entire. In S. cinerea and S. aurita the glands are much nearer, if not on the margin, and are often at the apex of teeth. The position, therefore, of the glands and nature of the margin seems to afford a frequent (though not a constant) guide to the possible origin and position of some of the varieties of S. Smithiana, such as velutina and Serruginea. Var. velutina, which, in a greater degree than all the other varieties, appears to have arisen from the crossing of S. viminalis with S. cinerea, is said to be everywhere rare. Specimens which scem to be referable to it I have seen from Kelvington, N. York- shire (J. G. Baker), named S. cinerea-viminalis, Wimm., and well marked. Less well-marked are examples from Llangarren, Herefordshire (A. Ley); and Pendeford Mill, Staffordshire, named S. rugosa (Dr. Fraser). Dr. Fraser has also found. plants near Wolverhampton, which both Leefe and Boswell-Syme have referred to S. holosericea, Willd., Koch (=S. velutina, Schrad. sec. And.). These, which are g, are certainly rightly placed under or near velutina, but they are not very like Wimmer’s examples of S. holosericea, Willd. ; and the flowering-twigs and buds are more glabrous than Koch's comparison of them with those of S. cinerea REVISION OF THE BRITISH WILLOWS. 419 would imply. Koch says that the catkins, buds, and twigs of holosericea bear so great a resemblance to those of cinerea, that it. is impossible to separate the two plants except by the leaves. Some of the plants named Salix rugosa, Leefe, seem to properly belong to velutina. ^ Leefe's own specimens (Sal. Brit. Ess. No. 30, and Sal. Exs. i. No. 22)—of which he writes, “I find it difficult to say whether this plant is nearer to S. holosericea, W., or S. Smithiana, W.; with much hesitation I have thought it on the whole to belong to the latter"—are very like Wimmer's example (Coll. Sal. No. 100) of holosericea, Willd., except that Wimmer's is more pubescent. (Andersson remarked of the Sal. Brit. Ers. No. 80 specimens that they seemed to him to be S. acuminata, Sm.; but Leefe justly says that they do not at all resemble S. acuminata, Sm. Eng. Bot., and Ward declares them to be true “ holosericea, Hook.") Between Dr. Fraser's holosericea and the plants from Pendeford Mill, named by him rugosa, I see no essential difference in the leaves. The latter are 9 plants, and from the tomentose twigs &c. seem to be velutina. So also is & “rugosa” from Clevedon (J. W. White); but other plants, from various sources, named rugosa pass by degrees into sericans f. subobscura. 8. ferruginea (Salix ferruginea, G. And.). This, which Andersson makes a variety of his y. velutina, resembles an intermixture of S. aurita and S. viminalis, though Whether it is really a hybrid of these species requires proof. Wimmer quotes English specimens (from Piuley, T. Kirk), named S. ferruginea, under his S. cinerea-viminalis ; and the examples I have seen from the same locality and collector seem probably referable to velutina, which, as has been already mentioned, Includes in part S. cinerea-viminalis, Wimm. Var. Jerruginea is normally a smaller plant in every respect than "LY of the foregoing varieties, and has more slender and more glabrous twigs, smaller leaves, usually broader above the middle and almost destitute of pubescence, and smaller catkins, An- ersson points out that ferruginea may be the same as S. obscura, Dill, a plant very difficult to distinguish from the smaller states of the modification subobscura of sericans, from which statement t may be gathered that that form is one of the links between erruginea and sericans. Boswell-Syme makes S. rugosa, Leefe, a Variety of Serruginea, distinguished by the different style and 420 - DR. F. BUCHANAN WHITE’S stigmas and more pubescent leaves—all very variable characters 1n this group. Salix rugosa, as said already, I prefer to place under velutina, though often passing into sericans. Some states of plants which I believe to be ferruginea very much resemble Wimmer's examples of S. holosericea, Willd. ; and just as sericans passes into velutina, so does the latter, I think, into ferruginea. Herbarium specimens named ferruginea are rather a mixed lot, including not only S. cinerea and velutina, but even perhaps sericans. Leefe’s ferruginea (Sal. Exs. iv. No. 89), received from Woburn, seems to be the same as N. J. Andersson's ferruginea; and Leefe says that it is G. Anderson’s plant, but not that of Eng. Bot. On the other hand, another ferruginea published by Leefe (Sal. Brit. Exs. No. 35, and Sal. Ess. iii. No. 63), brought from Essex, is rather a different plant. Leefe says that it is the same as Eng. Bot. Sup. t. 2665, but I cannot say that I see the resemblance. Andersson said of it that it was nearest to S. holosericea, Willd., and it seems to be very close to if not identical with velutina. . Many of the plants called ferruginea by English botanists have the leaves densely hairy below, and, apart from the other differences which they display, cannot well be the same as the plant defined by Borrer as minutely hairy. In a living and wild condition, ferruginea is known to me in one locality (in Perthshire) only. Here it has all the appearance of being a natural hybrid between S. viminalis and S. cinerea, since S. aurita does not occur in the immediate neigh- bourhood. I suspect that here, therefore, it is really a hybrid between viminalis and cinerea, with a greater proportion of the latter than of the former in it; aud that, since the common British form of S. cinerea is unlike the common Continental form, ferruginea has not been so readily recognized as a state 0 S. cinerea x S. viminalis. e. acuminata (Salix acuminata, Sm.). Wimmer, thinking that the name acuminata, having been variously applied, is ambiguous, replaced it by his own name Calodendron, which, on account of the doubtful origin of the plant, he prefers to his earlier Caprea-dasyclados. Andersson quotes S. Calodendron, Wimm., as a synonym of A acuminata, Sm., and says that the latter is elosely allied to, if not identical with, S. dasyclado, Wimm. (=s. longifolia, Host sec. Wimm.). REVISION OF THE BRITISH WILLOWS. 421 Comparing Leefe’s specimens (Sal. Brit. Exs. No. 37, and Sal. Exs. ii. No. 27) of Salix acuminata with Wimmer's dasyclados (Coll. No. 99) I find that they are practically identical. Wim- mer’s exponent (Coll. No. 100) of acuminata is also like Leefe's specimens, but with a somewhat greater look of Caprea. Of Leefe's No. 37, Andersson remarked that it was most certainly dasyclados, and very markedly distinct from acuminata, Sm. ; but Leefe maintained that it was the acuminata, Sm., Eug. Bot., and cites Borrer as confirming the name. From a capsule preserved with Smith's original drawings, I have no doubt that Leefe is right as regards his plant, and I have also no doubt that dasyclados is a synonym of acuminata. From the specimens which I have seen I am rather inclined to think that acuminata deserves a more distinct position than the other varieties of Smithiana; but as Andersson has placed it among them, I do not venture to remove it. Its large catkins, densely hairy with crisped hairs, and the long erect woolly pubescence of the capsules, make it easily recognizable. The stout twigs are also most generally densely tomentose, and the large leaves are usually broader upwards and glaucous below. Andersson defines three modifications—glabrescens (to which most of the British specimens belong), virescens, and cinerascens (also British). The parentage of acuminata is doubtful. Andersson, by saying that its characters are intermediate between S. viminalis and S. Caprea, seems to think that it has sprung from these species. But Wichura has proved by experiment that sericans is produced by the union of viminalis and Caprea ; and it is difficult to believe that in acuminata we have not a different combination, arising perhaps from a second hybridization with S. Caprea. From the dubiety which involves the parentage of most of the varieties of Smithiana, a series of experimenta in hybridizing S. viminalis with S. Caprea, S. cinerea, and S. aurita—both sepa- rately and in combination with the resulting hybrids—is much to be desired. Till this is done it seems hopeless to expect that the various forms can be extricated from the confusion in which they are at | resent. Group 8. NIVES. 12. SALIX LANATA, L. This handsom2 species see.ns to be a little more variable than 422 DR. F. BUCHANAN WHITE'S the British descriptions of the plant generally indieate. The leaves vary a good deal in size; in shape from suborbicular to oblong-obovate ; and in the amount of pubescence from being per- sistently woolly to nearly quite glabrous almost from the beginning. The catkins vary in size and in the eclour of the pubescence of the scales, which, though usually yellow, sometimes fading to grey, is occasionally grey from the first. ' Andersson describes several modifications, but there are none which deserve varietal rank. Salix lanata seems almost confined (in Britain) to Forfarshire and Aberdeenshire, or at any rate is most common in these counties. Whilst recorded long ago from Perthshire, it seems to have been lost sight of there till recent years, when it was redis- covered by Messrs. Brebner and Haggart and myself in a few places in the Glen Lochay hills. Var. Sadleri (Boswell-Syme). The plant described by Boswell-Syme as Salix Sadleri has now been in cultivation for several years, and has developed some features not seen in the original wild specimens. The stems have increased in thickness and become tolerably stout; the leavee have become larger (some being1}x1#inch). M any of them are more decidedly cordate at the base, and many are also furnished with large ovate-acute glandular-toothed stipules. Though the margins of the leaves were described as entire, yet even in the original specimens the edges of some of the leaves are, more especially towards the base, very finely glandular-serrate or crenate, the teeth, however, being almost reduced to glands. In mature leaves the margin is recurved or thickened. The catkins, which, so far as I have seen, have not increased much in size, may be described as either terminal or as terminal and lateral, according to the view taken of the structure of their peduncles. They are situated at the end of shoots which have two or three leaves. These shoots might be considered to be leafy peduncles, but as their leaves are furnished with stipules and have buds in their axils they are really perhaps a permanent part of the plaut, and in this case the peduncle of the catkin must be described as leafless. The scales are oblong and concave, with the tip rounded, emarginate or truncate ; in colour they are greenish, with the apex occasionally tinged with red or very shortly black-tipped, and are clothed with long white hairs, The greenish-yellow REVISION OF THE BRITISH WILLOWS. 423 style is about as long as the naked young ovary; the yellow stigmas are bifid and spreading, and half as long as the style ; the nectary and the pubescent pedicel of the ovary resemble those parts in Salix lanata. S. Sadleri has been supposed to be a hybrid of S. lanata and S. reticulata ; but after long and repeated study of both wild and eultivated specimens, I cannot recognize in it any trace of S. reticulata, nor of any species but S. lanata, a remarkable form of which I think it must be considered to be. Serrate leaves are not common in 5S. lanata, though Andersson mentions their occurrence, and I have seen examples. Moreover, in S. Sadleri the serration is obscure and not constant. The chief differences lie in the small catkins and in some parts of their struc- ture, but even in these the Janata element is preeminent. The whole facies of the plant is that of a small form of S. lanata. X SALIX suPERATA, n. hybr. (S. lanata x S. reticulata.) A willow which grows, in company with S. lanata and other mountain-species, on the rocks at the head of Allt Innis Choarach, Glen Lochay, Perthshire, has required a considerable amount of study to decipher. At first sight the facies of it does not distinctly suggest any particular affinities ; but on examination, and keeping im mind the species with which it is associated, a clear relationship to S. lanata is revealed. The other parentage seemed very difficult to guess, but a certain uudescribable appearance of the plant hinted that it might be found in S. reticulata. The possibility of this suspicion being correct further study has made a probability. The following description is taken from dried specimens :— A very low bush, with erect or ascending branches ; stems rather stout, twigs moderate. Bark rich brown, rather shining ; young shoots greenish, sparingly hairy, but soon glabrous. Buds shortly oval-oblong, obtuse, at first greenish and very slightly hairy, then pale brown and glabrous. Leaves more or less obovate, base attenuate and often unequal, apex shortly pointed and plicate- twisted ; leaves at the top of the shoots the largest (average size about 14 x 2 of an inch), those below smaller and proportionately narrower; upper surface yellowish green, scarcely shining, flat, with slightly impressed veins; lower subglaucous dull, with the chief veins conspicuously raised and reticulate; upper surface slightly woolly with short hairs, under with long and short hairs, 424. DR. F. BUCHANAN WHITE'S both soon nearly glabrous; margin slightly cartilaginously thick- ened, entire or with a few minute glandular teeth, especially towards the base; petiole medium in length, channelled above, more or less tinged with purple, which colour sometimes also spreads to the midrib, veins, and margin, though these are more frequently yellowish ; stipules, when present, rather small, ovate, glandular-toothed. Catkins (d) terminal on the branches, sessile, short but stout (4 x inch) ; scales roundish, spathulate, rather broad, thin, brown with darker veins, densely pubescent on both surfaces with long straight white hairs; filaments glabrous; nectary half surroundiug the base of the filaments, short, split into 2 or 3 obcuneate pieces. From Salix lanata the more erect growth, shape of the leaves, and sessile woolly catkins seem to have been derived; whilst to S. reticulata the smaller size and glabrosity of the leaves, the brighter green of the upper surface, the colour of the petioles, veins, and margins, the rounded scales of the catkins, and above all the structure of the nectary, appear to be due. The elevated reticulation of the underside of the leaves is not greater than that which may often be seen in S. Janata, but the faint glandular toothing is very similar to what is shown by some specimens of S. reticulata. I have seen other plants which may alao prove to be hybrids of S. lanata and S. reticulata, but judgment on them must be reserved till a larger series of specimens has been obtained. X Sarx STEPHANIA, n. hybr. (S. lanata x S. herbacea.) Near Coire Dhubh Ghalair, Glen Lochay, Perthshire, Mr. D. A. Haggart found, two or three years ago, a very small willow whose affinities at first sight were rather difficult to determine, but which after much consideration I concluded to be with S. lanata and S. herbacea, in whose company it grows. More recently I have found in the same locality another form of the plant, and its cha- racteristics seem to point pretty clearly to a parentage from these species, It may be expedient to describe both forms. Though I have seen the original plant growing, the following description of it 18 taken mostly from dried specimens. Stems more or less buried ; branches short, moderately slender, and rather tortuous ; young shoots downy, but soon becoming -——— REVISION OF THE BRITISH WILLOWS. 425 glabrous. Leaves nearly orbicular, slightly cordate at the base, slightly longer than broad (the largest about Z inch long); margins more or less crenate-serrate; surfaces bright pale green, but not shining, slightly hairy with white woolly pubescence, more especially on the margin; veins pellucid, anastomosing, above impressed (when young) or slightly raised (when old), below ele- vated ; stipules minute, gland-like. Catkins (large for the size of the plant, being 3 inch in length) rather lax-flowered, on lateral peduncles which are furnished with one or two leaves; scales small, roundish, fuscous-brown but paler at the base, clothed with long white hairs; ovary lanceolate-conic, rather obtuse, glabrous except at the extreme base and on the very short pedicel (thus resembling some forms of S. herbacea); pedicel equal to or shorter than the nectary ; style moderately long, slender, purple (whieh colour occasionally tinges the ovary also); stigmas slender, bipartite. The more recently found form was described when living. It makes a low plant, with flexuous slender rooting stems buried amongst moss. Bark of the twigs rich but pale brown (becoming dark when dried) or greenish brown, glabrous or slightly downy ; young shoots greener and sparsely pubescent with long white hairs; buds rather long, oblong, blunt, at first green and hairy, then brown and glabrous. Leaves wavy and balf-folded, from orbieular to oblong in shape (the largest 14x14 inches); base cordate, rounded, or slightly attenuated; tip often shortly pointed and twisted; upper surface rather dark green and somewhat shining, lower paler and subopaque, both with numerous minute white dots, which disappear more or less in the dried plant; youngest leaves more or less clothed with long white hairs, but soon becoming glabrous; veins pellucid, reticulate, impressed above (or slightly raised in the older leaves when dried) and elevated below ; margin crenate-serrate, pinkish (as are the veins) in the youngest leaves (this colour disappears in the dried plant) ; stipules few, ovate, glandular-serrate. Catkins( the largest 1 inch long) terminal on leafy stalks ; scales spathulate and involvent, mostly pale and either scarcely or distinctly darker-coloured at the apex, clothed with long white hairs ; ovary subulate, glabr ous, on a pedicel as long as or slightly shorter than the long thin linear nectary ; style pale, long and slender ; stigmas long, bifid, spreading. Whilst in both of these forms the herbacea element is very per- 426 DR. F. BUCHANAN WHITE'S ceptible in all the parts, the derivation from Zanata is shown chiefly in the catkins, though it has left its impression in the other organs as well. I have some wiilows (from the same locality) under observation which may prove to be forms of Salix Stephania more nearly related to S. Janata than the two described above. In Juste Botan. Jahresber. for 1885 a bybrid willow found by Strömfelt (in Iceland P) is referred to S. herbacea x lanata, n. hybr. f. a. pubescens, Lundstr., and B. glabrata, Lundstr. 18. Sarix Lapronvum, L. Out of S. Lapponum, as in the case of many other willows, several supposed species were manufactured in the early times of salicology. Most of these have now been abandoned, though in the British lists three are still retained as varieties, viz. :— a. arenaria (L. ex p.), b. Stuartiana (Sm.), and c. pseudo-glauca, Syme (— S. glauca, Sm.). Regarding these, Boswell-Syme says that in his opinion they are scarcely distinguishable; J. D. Hooker characterizes a and b as only slight varieties, and doubts S. glauca, Sm., being the same species; and Babington thinks that S. glauca, Sm., is probably not a native, and that it is hardly the same as a and 5. These varieties are not admitted by the Continental sali- cologists, who, however, distinguish some other forms. Anders- son admits one only—Aelvetica (Vill.), whieh Wimmer considers to be a distinct species more allied to S. glauca, L., than to S. Lapponum. "Wimmer gives two varieties—b. marrubiifolia (Tausch), a very woolly form; and e. Daphneola (Tausch), a narrow-leafed glabrous or almost glabrous plant. Of the British varieties, a. arenaria and b. Stuartiana have no claim to be maintained as distinct. b differs from a only in the greater woolliness of the leaves, and is probably the same as Wimmer's var. b. marrubiifolia, which also cannot well be retained. As regards 8. glauca, Sm., the case is different, as will be explained in treating of the var. or subspecies helvetica (Vill.). But whilst no varietal names can well be given, it is not to be denied that S. Lapponum shows a considerable range of variation. The leaves of our Scottish plant vary both in shape and size and in the amount of pubescence. In shape they range from narrowly- lanceolate to roundly-ovate, some of the latter simulating the form REVISION OF THE BRITISH WILLOWS. 427 argentea or arenaria of Salix repens, whence arose Linné's con- fusion of the name arenaria. Whilst the leaves are usually well- clothed with pubescence, specimens occur which approach a gla- brous condition, but in the absence of fruit cannot be identified with Wimmer's var. Daphneola. The catkins vary a good deal in size, and the capsules both in shape and in the structure of the styles and stigmas. In flowers, which seem to be about the same age, and hence comparable, one set has lanceolate-subulate subacute capsules, distinctly pedicelled, but with the pedicei usually shorter than and very rarely as long as the nectary; whilst another set has ovate-conie smaller sessile capsules with longer styles. In other respects the plants do not present much difference, and intermediate forms connect the two seta. Though S. Lapponwm is most usually a truly alpine species, rarely descending (in Central Scotland) below an altitude of 2000 feet or thereabouts above sea-level, it does occasionally occur in the low ground. On the south side of the Ochil Hills, in Perthshire, a few bushes (discovered by my friend Mr. W. Martin) grow at the ed ge of a field at an elevation of only 700 feet above sea-level. So far as I know, S. Lapponum has not been found in any other part of that range of hills, which, moreover, are as regards that neighbourhood almost devoid of alpine plants. In this locality (which I have visited) it grows with S. pentandra, and could not, to all appearance, have been brought down by water from any sufficiently high altitude. As mentioned in the ‘Student’s Flora’ and elsewhere, S. Lap- ponun has also been found near Edinburgh. The recorded locality 18 Colinton (sometimes erroneously written Collingtou) Woods; and in Edinburgh University Herbarium is a specimen (with 9 catkins) collected there by Greville in 1824. More recently have seen, in the same herbarium, other specimens, collected many years ago, from Craigcrook and Dalkeith Woods, both of which places are near Edinburgh. Whether the species has any claim to be considered native in these three localities, or whether it still occurs there, local botanists must decide. In England this species has been found on Helvellyn only (B. King, 1880). - Lapponum hybridizes with several other species, and some of these hybrids have been found in Britain. Others probably Occur and should be looked for. 428 DR. F. BUCHANAN WHITE’S Var. or Subspecies Salix helvetica, Vill. Whilst Andersson considers S. helvetica, Vill., to be a variety of S. Lapponum, Wimmer treats it as a distinct species, placing it in his ninth tribe, whereas S. Lapponum is in the fifth. At the same time he says that it must be left to the Swiss botanists to decide whether it is a variety or hybrid of S. Lapponum or a dis- tinct species. S. helvetica combines to a certain degree the characters of S. glauca and S. Lapponum. From the latter it differs chiefly by the leaves being always or finally glabrous above, by the catkins being on distinct leafy peduncles, by the styles being bifid or subbifid, and by the paler scales. One or other of all these cha- racters may occasionally be seen to a certain extent in true S. Lapponum, but yet S. helvetica seems to merit varietal or sub- specific rank. From a comparison of specimens and descriptions, I had already come to the conclusion that Smith's S. glauca (which is not the Linnean species) was the same as S. helvetica, when I noticed that Walker-Arnott had apparently expressed what is practically the same opinion. As a synonym under his variety B of S. arenaria (=S. Lapponum) he mentions S. glauca, Sm., and in the notes he says: “ For our var. B we give no stations because we have no reason to believe it indigenous, ..... it is commonly cultivated . . . and is common in Switzerland, where we believe our var. a does not occur.” Though he is wrong in the latter opinion, yet S. helvetica is evidently the Swiss plant he had in view. Whilst, therefore, there is no doubt that Smith's S. glauca is the same as S. helvetica, Vill., it is not quite certain that it is a British plant. Smith gives as localities *in the Highlands of Scotland. Mr. Dickson. On the Clova mountains; Mr. G. and Mr. D. Don. Hooker." Walker-Arnott says that “Mr. Don's specimens now before us from the Clova Mountains are the same as S. arenaria, E. Bot. ;" and Babington states that Smith's specimens came from Mr. Crowe's garden. On the other hand, there is before me a specimen in Edin. Univ. Herbarium labelled by Winch “ Salix glauca, Ben Lawers.” This agrees with the plant cultivated by Mr. Leefe as $. glauca, Sm., “ re- ceived many years ago from the Cambridge Botanic Gardens,” and both are, without doubt, referable to S. helvetica. If Winch’s specimen really came from Ben Lawers (and I see no REVISION OF THE BRITISH WILLOWS. 429 reason why it should be doubted), then Salix helvetica must be admitted as a British plant. At the same time it is desirable that it should be rediscovered. As regards the figures of Smith's glauca, Wimmer condemns, on the authority of Tausch, Eng. Bot. t. 1810, as representing S. glauca, L., and though he cites Sal. Wob. t. 68 under that species, he remarks that it is not good, which, as it illustrates a form intermediate between S. glauca, L., and S. Lapponum, it could not well be expected to be. Smith cites Haller Hist. t. 14 as well representing his glauca, and Wimmer cites the same under S. helvetica, which affords further evidence of the identity of Smith's species. Smith himself, however, gives S. helvetica, Vill., as a synonym of S. arenaria, L. (— S. Lapponum). In Europe, whereas S. Lapponum is both arctic and alpine, S. helvetica is alpine only (Switzerland, Tyrol, and Dauphiné). X SALIX AURITA-LAPPONUM, Wimm. In Edinburgh University Herbarium is an old specimen of a willow collected, I think, by the late Prof. J. H. Balfour, and labelled o Colinton, Edin.,” but without a date. It has one 9 catkin, and is undoubtedly a hybrid between S. Lapponum and S. aurita, both of which species occur, or used to occur, at Colinton, although, as mentioned under that species, it is doubtful whether S. Lapponum is native there. Evenif it were planted, that is no reason either for or against the hybrid being of spontaneous origin. There is also a leaf-specimen, collected near Craigerook In 1832 by the same botanist, which, from the structure of the leaves, I refer to this hybrid. As stated under S. Lapponum, that species has also been found at Craigcrook. Andersson combines the hybrids of S. Lapponum with A Caprea, S. cinerea, and S. aurita, under the name S. Lesfadiana, Hartm., and places S. aurita-Lapponum, Wimm., as b. opaca, 2° subaurita, Like other hybrids, S. awrita-Lapponum exhibits various com- binations of the characters of its parents, but is best known from Its close ally S. cinerea-limosa, Lestad., by the rugosity of the leaves, especially the young ones. It occurs in several parts of Europe (Lapland, Sweden, Silesia, Switzerland), and ought to be found in the Scottish Highlands, where the parent species not unfrequently grow in sufficient proximity. LINN. JOURN.— BOTANY, VOL. XXVII. 26 430 DR. F. BUCHANAN WHITE'S A willow, in Kew Herbarium, collected by Lightfoot, but without locality, date, or name, seems to be Salix aurita-Lap- ponum. X SALIX CINEREA-LIMOSA, Lestad. (S. Lapponum x 8 cinerea.) A specimen, in Edinburgh University Herbarium, labelled “Salix cinerea, Carlowrie, 1838” by, I think, J. H. Balfour, is evidently a hybrid of S. Lapponum, and is very like S. aurita- Lapponum. From the densely black-pubescent twigs, the abundant rusty cinerea-like hairs on the underside of the leaves, and the absence of the rugosity of aurita, it is most probably a hybrid with S. cinerea rather than with S. aurita. The specimen has numerous 9 catkins, but only young leaves. Carlowrie is near Edinburgh, and though there is no record of S. Lapponum having been found, it may have been planted there as well as in the other places in the district. The hybrid itself is not at all likely to have been introduced, but rather to have had a spontaneous origin. S. cinerea-limosa, Liestad. (S. cinerea- Lapponum, Wimm. ; 8. Lestadiana, Hartm., 9. opaca, 1° subcinerea, And.), seems to be of rare occurrence, recorded from North Scandinavia only. It is not so likely to occur on our hills as S. aurita- Lapponum. X SALIX SPURIA (Schleich.), Willd. (S. Lapponum x 8. Arbus- cula.) Since S. Lapponum and 8. Arbuscuca are (with the exception of 5. herbacea) probably the two commonest of our alpine willows, they, as might be expected, eross with each other, though, on the Coutinent at least, apparently not so frequently as might be supposed. ‘Tue hybrid, though at first at least not recognized as such, has been known for a considerable time under the name of S. spuria, Schleich., Willd.; but Wimmer, according to his custom, altered the name to S. Lapponum-Arbuscula. It occurs both in Switzerland and the Tyrol, but, judging by its absence from many herbariums, seems to be rare. In Britain, though it 18 about Seventy years since it was first gathered, it has nol, until quite recently, been identified, having usually been passed over as a form of S. Arbuscula. REVISION OF THE BRITISH WILLOWS. 431 Although I have not been able* to examine European examples of Salix spuria, I have before me a series of willows which are beyond doubt hybrids between S. Lapponum and S. Arbuscula, and which must therefore be called S. spuria. Though, as in the case of other hybrids, some specimens show more affinity with one parent than with the other, the examples I have seen are all tolerably intermediate in character. Judging from the words of Andersson and Wimmer, and from the names given to it by Seringe and Gaudin, the Continental plant seems to have a greater resemblance to S. Lapponum than has the British one. Andersson says that it is so similar to that species that it can scarcely be distinguished, except by the shorter and more compact catkins surrounded by a few leaves, but especially by the much harder and more glabrous leaves finely serrated on the margin. Wimmer describes it as approaching S. Lapponum in the size and clothing of the leaves, and shape and length of the catkins; but to 9. Arbuscula in the glandular-serrate leaves, Which are somewhat shining above, in the narrower ferruginous scales, and in the structure of the stigmas. The Scottish plant, on the other hand, is more likely to be taken—as indeed it has been—for S. Arbuscula than for S. Lapponwm. In size it is nearer the former than the latter, but otherwise it shows a combination of the characters of both. From S. Arbuscula it may be distinguished generally by the duller colour of the leaves, which are more or less, but never excessively, pubescent—the pubescence combining the silkiness of Arbuscula with the woolliness of Lapponum, by the finer and more scanty serration of the leaf-margins, by the longer shape of the capsules, longer styles, and usually narrower scales darker at their tips; and from $. Lapponum by the firmer and more shining leaves, which are more nearly glabrous and have more or less serrate margins, by the smaller catkins with short leafy Peduncles, and by the short stigmas. The characters, however, are more readily seen than described. I have seen British specimens from the Breadalbane Moun- tains, In the former it seems to have been first gathered by Sir . D. Hooker, since in Mr. Hanbury's “ Boswell Herbarium " is * Since this was written I have received (through the kind services of Mr. A. Penn ett) Scandinavian specimens of Lapponum x Arbuscula, which prove to be tentical with some of our Scottish examples. 262 482 DR. F. BUCHANAN WHITE’S a specimen (without date and without flowers) labelled S. Arbus- cula. Of recent years it has been found in several places in Breadalbane, with both d and 9 catkins, by Messrs. Meldrum, Haggart, and Brebner, and by myself. Group 9. Nrrtnpts 14. SALIX Myrsinrres, L. Considering how, on the whole, well-marked this species is, great discrepancies occur in the various descriptions. Andersson says that the ovary is sessile, with the nectary passing its base; Wimmer (with whom Grenier agrees) that it is shortly pedi- cellate with the nectary reaching its body ; W. J. Hooker that it is sessile (in S. procumbens nearly sessile) ; Walker-Arnott (for both “ species") that the pedicel is usually as long as or at length longer than the nectary ; Boswell-Syme that the pedicel is about as long as the nectary ; Babington that the ovary is sub- sessile; and J. D. Hooker that it is distinctly pedicelled. Then as regards the shape of the ovary, Andersson describes it as ovate-conic rostrate (in B. Jacquiniana, thickly ovate and scarcely pedicellate) ; Wimmer, conic-oblong, always obtuse 12 a, more slender and often somewhat acute in B. Jacquiniana; Koch, lanceolate-acuminate from an ovate base, at first sessile and hairy, at length very shortly pedicellate and glabrous; Grenier, ovoid shortly conic; W. J. Hooker and Walker-Arnott, lan- ceolate; Boswell-Syme, lanceolate-conieal or conical-subulate; Babington, ovate-subulate. (By Andersson the d catkins are described as yellowish, the anthers at length becoming blackish; but Wimmer more correctly says that the filaments are purplish, and the anthers purple- violet.) From the descriptions of the ovary, it would appear that British botanists attribute a longer pedicel to the species than the Continental botanists do, and that on the whole they are inclined to describe the shape of the ovary as lanceolate. The result of the examination of a number of specimens shows that there is a considerable range of variation in the lengtb of the pedicel and shape of the ovary, without, however, any notably extreme forms. The Central-European plant seems usually to have a shorter capsule and pedicel than the North-European. Seottish examples are more (though not altogether) in accord- REVISION OF THE BRITISH WILLOWS. 433 ance with the latter than with the former. In them the pedicel is always (so far as I have seen) present, and is often twice as long as the nectary, though it varies in length even in the same catkin. In addition to these variations in the ovary, the structure of the style and stigmas, the shape and size of the catkins and of the leaves, as well as the habit of the plant, are all subject to modification, so that as regards the varieties serrata and arbuti- Jolia of British lists there is no reason for which they deserve to be retained. Nor is it by any means evident that Salix procumbens, Forbes (now considered by most botanists to be a form of S. Myrsinites) has any characters to warrant its retention. Walker-Arnott maintained it as a distinct species chiefly on account of its elongate catkins ; but, as Boswell-Syme points out, this is not à character of any constancy in S. Myrsinites, nor are there any sufficiently distinct characteristics in the habit of the plant, the shape of the leaves, or the structure of the style. x SALIX WanrENBERGIL, And. (S. Myrsinites x S. nigricans.) In the * Prodromus! Andersson uses the name S. myrsinitoides, Fr., for the willow which Wimmer calls S. Myrsinites-nigricans ; but in Blytt's «N orges Flora,’ pt. ii. 1874, he alters the name to 8. Wahlenbergii, And., and gives as synonyms S. punctata, Wahl., 8. myrsinitoides, Fr., S. nigricans *borealis B. punctata, Hartm., and 8 Myrsinites-nigricans, Wimm. o. , Both Andersson (in the * Prodromus’) and Wimmer (in his Salices ’) appear not to have been very well acquainted with the Species, as their descriptions do not altogether agree, and neither was sure of the distribution. From the ‘ Norges Flora,’ however, it would seem that Andersson had become more familiar with the plant, since the description is greatly amplified. After remarking that, through the hybridization of the very variable Myrsinites With the still more variable nigricans, a great number of intermediate forms occur, he proceeds to notice three chief modifications, namely, a. subnigricans, b. coriacea, and e. sub- Myrsinites, Since, however, these forms pass imperceptibly into each other, it seems scarcely worth while retaining these names as those of distinct varieties. In its best form S. Wahlenbergii combines the characteristics of its Parents, deriving from Myrsinites the rigidity, glossiness, 434 DR. F. BUCHANAN WHITE'S and in part the venation of the leaves, the often erect leafy- peduncled catkins, and the structure and colour of the style and stigmas; and from nigricans the somewhat tomentose twigs and leaves, the greater thinness of the latter, and their greater tendency to become black in drying, the often longer petioles, and the often longer pedicels of the capsules. But, as Andersson remarks, some specimens are very difficult to separate from nigricans, and others pass imperceptibly into Myrsinttes, and thus it is often by no means easy to determine some of the less intermediate examples. In Britain Salix Wahlenbergii is possibly not very much rarer than S. Myrsinites itself, since it occurs in many of the localities of the latter, though it has been passed over as Myrsinites and sometimes as nigricans. I have seen good intermediate speci- mens from Perthshire, Forfarshire, and Aberdeenshire—first gathered in the latter two counties by J. H. Balfour, Greville, and their contemporaries ; and in addition many examples gradually passing into either Myrsinites or nigricans, and often very difficult to separate. Neither Wimmer nor Andersson seem to have been acquainted with the d plant, though the former quotes of the catkins “ d elongati, graciles." The g has, however, been found on Ben Laoigh by the Messrs. Groves, and on Ben Heasgarnich by myself. These g plants have leaves near Myrsinites, catkins short, thick, on leafy peduncles, and anthers subglobose and yellow, this yellow colour being the chief point in which they differ from Myrsinites. ; Salix MacNabiana, M acGillivray (‘ Edinb. New Philos. Journ. ix. p. 335, 1830), is possibly this hybrid, but MacGillivray's own herbarium specimens (now in the possession of Dr. Roy, of Aberdeen)—from the Corrie of Loch Kandor, where the hybrid is common—are very close to, if not identical with, Myrsinites, though they are too poor for absolute certainty. A hybrid between $. Myrsinites and S. phylicifolia probably occurs in Britain, but I have not yet seen exampies sufficiently certain to enable me to include it in the list. X SALIX SAXETANA, n. hybr. ($. Myrsinites x S. aurita.) A willow gathered by myself some years ago, and more recently by Messrs. Groves, on Ben Laoigh, Perthshire, puzzled REVISION OF THE BRITISH WILLOWS. 435 me for a long time. I have found quite lately, however, several bushes which, by their range of variation, indicate pretty clearly the affinities. The plant seems, with little doubt, to be a hybrid of Salix Myrsinites and S. aurita, both of which occur on Ben Laoigh, but is so altered that at first sight it does not sugges either of the parents. As yet I have seen 9 catkins only—from at least three bushes, but I think that the d also occurs *. The following description is made from a form tolerably inter- mediate in its characters. It makes a low thick-stemmed bush, with olive-green (turning brown when dried) slightly shining twigs, which are at first somewhat pubescent, but soon become glabrous. Leaves rather large (2 x 1 inches), especially at the top of the shoots, obovate, with a more or less oblique short point ; margin wavy, coarsely erenate-serrate ; upper surface dark shining green, with (when living) the veins slightly im- pressed ; under surface dull pale green or subglaucous, with the primary veins raised; young leaves slightly pubescent, soon becoming glabrous, except on (to a greater or less extent) the midrib and rather long petiole. Stipules very small, 1-cordate, glandular. Catkins moderate in size, erect or spreading, on long or short peduncles which are furnished with several small leaves Whose axils have buds in them; scales narrow, pointed, with the upper half or third fuscous black ; capsule rather small, conical- subulate, clothed with dense white hairs, and on a pedicel which is three to four times as long as the small quadrate yellow néctary; style rather short, bifid to, or nearly to, the base ; stigmas rather short, bifid, spreading. Specimens from another bush show much the same characters, * Since this was written, the supposed ¢ plant has flowered in cultivation. The following description was taken from a living specimen :—Catkins ovate- oblong, lateral, on a peduncle with about three leaves which are as long as the catkin. Peduncular leaves minutely stipulate, glabrous, pale dull green, with the chief veins yellowish green and impressed above ; below clothed with rather straight hairs on the midrib, chief veins, and apical margin, dull paler green, with the principal veins yellowish and raised ; veins translucent, those Which arise from the midrib becoming subparallel to it; margin finely glandular- serrate. Scales broad, greenish white below, the upper third or fourth becoming black (the scales at the top of the catkin are red between the pale and black Portions) ; upper part sparingly hairy with long whitish hairs; apex usually emarginate or erose, Filaments pale, glabrous. Anthers (with cells unequal) subquadrate, at first tinged with red, becoming yellow. Nectary rather small, thick, oblong, usually entire, but sometimes divided into two or more irregular Pleces, green or yellowish-green, The date of flowering was the end of May. 436 DR. F. BUCHANAN WHITE’S but the capsule is less subulate and more cylindrical and obtuse, and seems to show below the hairs a trace of the scaly pubescence which is sometimes characteristic of Salix Myrsinites. A third bush, while practically the same as the others, shows rather a greater affinity with S. aurita. The twigs are more slender ; the young leaves more pubescent; the margins of the leaves more serrate; the catkins smaller, with narrower, more ferruginous scales; the capsule more like that of aurita, with usually an almost obsolete undivided style, and very short, bifid, and erect stigmas. It will be seen from the foregoing that, though the leaves have to a great extent retained the outline of those of S. aurita, they have lost the rugosity characteristic of that species, though in their young state they show in part the pubescence. From S. Myrsinites the shining upper surface has been derived, but the substance is thinner than in that plant. The capsule struc- ture shows varying affinities with both species. Besides the Ben Laoigh examples, I have seen specimens collected in Clova, Forfarshire, by the Rev. W. R. Linton, which—although there are no catkins—I believe to be another form of S. saxetana. In shape the rather small leaves recall both S. Myrsinites and S. aurita. They are roundish- or oblong- obovate, with short more or less oblique points; green but (at least when dried) only slightly shining on eithersurface; margins fnely erenate-serrate; both surfaces sparingly pubescent—the pubescence becoming scantier, but scarcely altogether vanishing in the older leaves ; young leaves rugose, with impressed veins above and raised veins below; older leaves flatter above, but with conspicuous pale (when dried) raised veins below; petioles rather long ; stipules present, but small. I think that, so far as the leaves go, there can be no doubt as to the parentage of this plant. X SALIX SERTA, n. hybr. (S. M. yrsinites x S. Arbuscula.) A specimen in Mr. F. J. Hanbury's ** Boswell Herbarium," labelled “ Salix Arbuscula. Breadalbane Mts., Lyon,” seems almost certainly a hybrid between Myrsinites and Arbuscula. The leaves recall both species, and are rather small, slightly obovate, shining above, and dull and glaucous below, finely crenate-serrate on the margins, quite glabrous (except when very young, when they have the pubescence of Arbuscula), and veined REVISION OF THE BRITISH WILLOWS. 497 in the style of Myrsinites rather than Arbuscula. The catkins (9), which are borne on terminal and lateral, rather long, leafy peduncles, are 14-2 inches long, stout and cylindrical; scales small and rather narrow, very hairy ; capsules sessile or nearly so, oblong-conical, obtuse, coloured like those of Arbuscula, and hairy ; style slender, rather short, with the apex bifid; stigmas short, slender, bifid. The plant is quite intermediate between its supposed parents. Here also may belong a scrap in the same herbarium labelled " Salix prunifolia. Breadalbane Mts. J.D. Hooker.” If it does, itis nearer Arbuseula. The small leaves are rather intermediate ; the capsule more slender than in the above; the style very short, and the stigmas very short and thicker. Under S. Arbuscula, Walker-Arnott describes a var. B from Ben Lawers, with *leaves (broadly or roundish ovate, promi- nently veined above) green, but scarcely shining on both sides." Of it, he says that it is precisely intermediate between Arbuscula and Myrsinites, and may perhaps be a hybrid. It was found only once. 15. SALIX HERBACEA, L. S. herbacea varies in many small particulars, such as size, shape and extent of serration of the leaves, downiness of the young shoots, Ze, The pedicel of the capsule varies a little in length, and in being glabrous or silky, the silkiness sometimes extending to the base of the capsule. The style also varies in length. Occasionally the capsule has a few hairs on it, and more rarely well-developed lines of hairs. The most extreme state of this form which I have seen is in a plant, collected by Dr. Greville m Corrie Kandor in 1830, which has, moreover, the style bifid to the base. A less extreme form I have found near Glen Tilt. Andersson describes a var. subpolaris whose capsule has lines of hairs, but it differs from the type in also having entire leaves. "BAL Grauami (Borr.), Baker. (S. herbacea X S. phylici- Jolia?) Under 8, Myrsinites Sir J. D. Hooker has the following note:—“ 8 Grahami, Borr. MS., is only known from Q speci- mens, cultivated in the Edin. Bot. Garden, said to have been brought by Prof. Graham from Frouvyn in Sutherlandshire. 438 DR. T. BUCHANAN WHITE'S It appears to me to be a form of Saliz Myrsinites, with smaller catkins, paler scales, and a perfectly glabrous capsule with a rather long very silky pedicel; and not allied to S. polaris or herbacea. Syme suggests it to be a hybrid between herbacea and nigricans or phylicifolia; and Nyman a subsp. of S. retusa, L. The Engl. Bot. figures of the ovary and scale are very incorrect. A similar plant occurs in Muckish Mt., Donegal.” Having examined living specimens of both the Scottish and Irish plants, I think that Boswell-Syme is probably correct in his suggestion of the parentage; but I cannot agree with Leefe in considering that the Irish plant “is as nearly as may be identical” with the Sutherland form. In fact, I suspect (although, till it is proved by experiment, it is only a conjecture) that whilst S. herbacea is one of the parents of both, S. phylici- Jolia is the other parent of the Scottish plant, and S. nigricans of the Irish. If this be the case, the former must bear the name of S. Grahami, and the latter that of S. Moorei. The published descriptions of neither of them are absolutely correct, but it will be sufficient to indicate the points in which the two differ. X S. Grahami (Borr.), Baker. Young leaves brighter and more shinin g, rather broader, less pubescent, the pubescence being more of the nature of that of S. phylicifolia, but not markedly so. Scales involvent, broader upwards, obtuse, subemarginate at apex, hairy at the base and ciliate on the margin. Pedicel of the ovary silky-hairy, the pubescence spreading more or less over the base of the ovary, Which is otherwise glabrous. Style rather stouter ; nectary oval- oblong. X S. Moorei (* Watson, L. C.”). Young leaves duller, narrower, and more hairy, the pubescence resembling that of S. nigricans. Scales much longer in pro- portion, narrow oblong, subobtuse at apex, and more hairy. Pedicel of ovary apparently longer, glabrous or slightly hairy; ovary more or less pubescent towards the apex. Nectary linear- oblong, about as long as the pedicel, finally much shorter. The most apparent differences between the two lie in the very differently shaped scales and in the pubescence of the ovary. Boswell-Syme describes the peduncle of the catkin of the Suther- REVISION OF THE BRITISH WILLOWS. 439 land plant as glabrous, but, in all the specimens I have seen, it, as well as the rhachis, is distinctly pubescent. Stipules were unknown him and are certainly rare, but when present are small, more or less narrowly ovate, and toothed. Possibly they are more fre- quent in Salix Moore? than in S. Grahami. In some wild specimens of S. Moore? (kindly lent me by Mr. F. Moore) there are ripe capsules. These are subulate from an ovate base, and about } inch long. The facies of the wild specimens favours the theory that nigricans is one of the parents. The late Dr. Moore, in recording the occurrence of the Irish plant, says that it is the same as specimens of S. Grahami from the Sow of Athole in Perthshire. I have not been able to learn anything about this Sow of Athole plant; but in Borrer’s Her- barium at Kew is a plant, placed under S. Arbuscula, which was collected on that hill by Mr. J. Ball. Its condition is bad, but it seems not to be S. Arbuscula, and may be a form of the Sutherland Grahami, though not agreeing with it in the scales or leaves. X SALIX sIMULATRIX, n. hybr. (S. herbacea x S. Arbuscula.) Under this name I place specimens of four plants from the Breadalbane Mountains, which, though unlike each other, seem to be probably hybrids of S. herbacea with 5S. Arbuscula. They all however, require further investigation. These plants and their characters are as follows :— l. From Coire Dhubh Ghalair (J.Brebner). Nearer S. Arbus- cula than S. herbacea. From the latter it derives its habit, slender arcuate branches, roundish oval thinner leaves, and pseudo-terminal catkins (i. e. at the end of a branch and sub- tended by a leaf, but with a bud between the leaf and the peduncle); from S. Arbuscula it has the thicker trunk and more oval leaves, dull glaucous green below and with the smaller Veins less prominent. Whilst bearing a strong resemblance to Arbuscula, it is not exactly like any of the numerous specimens which I have seen. Compared with examples of the latter of the same age, it may be distinguished by the leaves being thinner, more shining above, and roundish oval in shape; by the habit and slender branches ; and by the position of the catkins. There 18 only one eatkin on the specimen, and that not in good con- dition. The capsules, whilst resembling those of Arbuscula, 440 DR. F. BUCHANAN WHITE’S seem to be a little longer in proportion to their size than in that species. 2. A specimen in F. J. Hanbury’s “ Boswell Herbarium,” on a sheet with the label “ Salix prunifolia, Breadalbane Mts., J. D. Hooker.” This has no flowers, but is in habit nearer herbacea than No. 1, with which in the leaves it quite agrees. 3. Meall Dhuin Croisg (W. Barclay and R. H. Meldrum). In habit and leaves near herbacea, but in its catkins nearer, appa- rently, Arbuscula. I once thought that this might be a form of S. Moorei nearer S. herbacea, but now I suspect that S. Arbuscula is more likely to be one of its parents. The catkins are lateral on leafy peduncles, and are moderately long. The capsules are pubescent, intermediate in shape between those of Arbuscula and herbacea, and pedicellate, with the pedicel about as long as the long linear nectary. The style is of medium length, and the stigmas rather short, thick, and cleft. The leaves much resemble those of herbacea, but are not quite identical. It is beyond doubt a hybrid of herbacea, but whether with Arbuscula is a little uncertain. 4. A specimen in the Edinburgh University Herbarium, labelled “ Salix vaccinifolia, Craig Chailleach, Perthshire; Dr. Hooker.” This has young catkins and young leaves only. The leaves seem essentially the same as those of Nos. 1 and 2. The catkins are lateral and terminal, on leafy peduncles, very small and subglobose. The scales are glabrous on the back and ciliate on the margins, and very similar to those of Aerbacea. The ovaries are like those of Arbuscula, but the style and stigmas like herbacea. Larger and older specimens are desirable. X SALIX SOBRINA, n. hybr. (S. herbacea x S. Lapponum.) A morsel of a willow gathered on the east side of Ben Chat, in Athole Forest, Perthshire, by Dr. Roy, of Aberdeen, and by myself, had long been a puzzle to me (especially as it had no catkins) till I came to examine a plant found by the Rev. E. S. Marshall in Glen Fiagh, Clova, Forfarshire, in 1888. These latter specimens, being in better condition and having 9 catkins, show that it is a hybrid between S. herbacea and S. Lapponum. The examples which I have seen are so intermediate between the parent species, that they are not strikingly like either of them. In size the plant is like herbacea, but in habit more like Lapponum, especially in the comparatively (to the size) stout, REVISION OF THE BRITISH WILLOWS. 441 knotty, tortuous, reddish-brown glabrous branches. The leaves are small, oval, at first woolly on each side, but becoming sub- glabrous, and slightly shining on each surface ; margin obscurely serrate; base rounded; tip (sometimes twisted) subacute and subcartilaginous, as in Lapponum; veins and margin pellucid ; veins finally raised and reticulate on the upper surface; lower surface lineately veined and reticulate with raised veins. Catkins short, few-flowered, lateral on long leafy peduncles ; scales large and broad, obovate, subtruncate, brown with darker tips and long white hairs; capsules sessile, white-woolly, at first conical, then subulate-conical; style moderate, about as long as the rather thick entire or bifid suberect stigmas; nectary long linear. Young shoots pubescent. A willow from Clova (Dr. Greville, 1824), in Edinburgh University Herbarium, may belong here, but if it does it is nearer herbacea than the above. It has leaves only. Since this description was written I have received specimens of A Lapponum-herbacea (no authority for the name, and S. alpestris, And., given as a synonym) from Sweden. These are, to all intents, the same as the Scottish plant, though rather more luxuriant, « Alpestris" is the name adopted by Andersson (in the * Norges Flora’) for the hybrid between 5$. herbacea and S. glauca, and is placed with four other, variously composed, hybrids under 5. norvegica (Fr.), Aud. I have also seen a more numerous series of specimens col- lected in Clova by the Messrs. Linton, as well as what appears to be another form (collected by the same botanists) from Craig- na-dala Beg, Aberdeenshire. These specimens exhibit various degrees of combination between the supposed parents. X BALIX MARGARITA, n. hybr. (S. herbacea X S. aurita.) In Messrs. Groves’ herbarium is a very curious willow, found by them on Craig Loigste, on the south side of Ben Challum, in Perthshire, in 1855; and growing in the Edinburgh Botanic Garden are two willows (for specimens of which I am indebted to Mr. Lindsay, the Superintendent) which, though not identical with the Messrs. Groves’ plant, are apparently another form of the same species. The Edinburgh Garden plants were found in 1875, by the late Prof. Dickson and the late Mr. J. Sadler, near ly ndrum, and probably both on Ben Challum, since a specimen 442 DR. F. BUCHANAN WHITE's sent by Mr. Sadler to Mr. Leefe, and now in Kew Herbarium, has Ben Challum given as the locality. Mr. Sadler describes the wild plant as having “a dwarf procumbent habit;" and Messrs. Groves’ specimens show this, but the cultivated specimens sent to me look as if they had become more upright. The following is a description taken from living specimens of ihe plant found by Mr. Sadler, and now in cultivation :— Twigs divaricate, slender, straight or subflexuose, purple-brown, glabrous, shining; older bark brown or very dark olive-green, roughened but shining; youagest shoots with white pubescence, when older green, or where exposed to the light dark red-brown or purplish, shining, but with a few hairs. Buds at first pubescent, then glabrous, yellowish-red, acute. Stipules (rare) half-cordate, hairy, glandular-serrate on the margin. Leaves thin, dark green, slightly shining above, paler and dull below, at first densely woolly with brownish-white hairs, at length almost or quite glabrous; roundish; base rounded, subcordate, or unequal and slightly cuneate; tip with a twisted point, or truncate, or subemarginaie ; upper surface concave ; margin thickened and incurved serrate-crenate, with incurved glandular teeth ; upper surface rugose from the impressed veins, under surface with raised reticulate veins, smaller veins pellucid ; petioles rather long and slender; the largest leaves li inch long and broad, but most of them much smaller; vernation involute. Catkins ( 9) small (less than 4 inch long), dense-flowered, on lateral leafy peduncles of about the same length as the catkin; peduncle-leaves 2-4, ciliate, with buds in their axils, and stipulate ; peduncle and rachis downy; scales oblong, narrow and long, concave, glabrous on the back, ciliate at the apex, and with a few hairs on the inner surface, greenish yellow, those at the top of the catkin tinged with pink at their tips; ovary conical from an ovate base, sub- obtuse, with coarse woolly white pubescence ; pedicel downy, nearly three times as long as the nectary, which is thin in texture, oblong or widened upwards, entire or cleft once or twice at the apex or nearly to the base; style very short, thick, greenish yellow; stigmas bifid, suberect, as long as the style. In one specimen the uppermost scales are widened upwards, broader and involvent, and have a few hairs on the back. In Prof. Dickson’s plant the characters are much the same. The older bark is duller, the buds are less acute, and the catkin- REVISION OF THE BRITISH WILLOWS. 443 scales more uniformly narrow, the catkins rather larger, and the stigmas spreading. Messrs. Groves’ specimens have often obovate and sometimes oblong leaves, the largest about 1 inch long by ? inch broad, but most of them less than half that size; slender, more flexuous twigs; young leaves less hairy; petioles rather shorter; the solitary catkin more slender, with scales shorter, broadly obovate, and involvent, the pedicel of the ovary rather shorter, and the stigmas larger and spreading. There can, I think, be no doubt that this species is a hybrid between Salix herbacea and S. aurita. The Edinburgh Botanic Garden plants are, in the shape of the scales and in the capsules, nearer aurita than herbacea, to which, in these parts, Messrs. Groves’ specimens are more related. Both sets are, however, intermediate in character between their parents. 16. SALIX RETICULATA, L. In several small points S. reticulata departs from the usual descriptions of this well-marked species, but these variations require no special notice. The margin of the leaf is described as entire, but on the appareut margin (though not usually on the real edge) there is frequently a row of glands which give the appearance of a minute serration. Andersson mentions two varieties—a. typica and p. nivalis (S. nivalis, Hooker), the latter being a remarkable small form which occurs in Iceland, Spitzbergen, and N. America. The Var. a. typica is divided into :—1. glabra (leaves quite glabrous on each side), and 2. sericea (leaves more or less villous, and the margin towards the base here and there glandularly subserrate). The latter is the more frequent form in limestone districts of S. Europe. Whilst the Scottish plant must be referred to l. labra, I think that the Glen Callater examples retain the hairs on the underside of the leaf (both surfaces are more or less hairy when young) for a longer period than the Perthshire ones do. I have seen two Scottish plants which have been referred to S. reticulata, but which seem to be evidently hybrids of it with other species, They are well worth further investigation. _ One of these plants is in the British Museum Herbarium, and 18 labelled « Salicis reticulate varietas ?, Ben Lawers, Perthshire, 1793, R. Brown.” Unfortunately it has no flowers, but it may be thus described :— 444 DR. F. BUCHANAN WHITE'S In habit like S. reticulata, but stems more slender, rooting, chestnut-brown. Petioles very short, being little, if at all, longer than the bud in the axil. Leaves narrowly oval or elliptic, some of them rather broader beyond the middle, attenuate at the base, flat above and below and glaucous below; midrib and veins slender, primary veins leaving the midrib at a very acute angle and running parallel to it, secondary veins anastomosing ; margin cartilaginously thickened, entire; texture apparently thinner than in reticulata; young leaves with a few hairs below, soon becoming quite glabrous. Largest leaves nearly 2 inches long by 2 inch broad; average size 11x 1 inch. The very short petioles &c. indicate the distinctness of this plant (which may be provisionally called S. sejuncía) from S. reticulata. The other plant referred to is in the Kew Herbarium, and is labelled “ Salix reticulata, L., And." (meaning that Andersson certified it), * Scotland, Herb. Lambert." It is quite a different looking plant from S. reticulata. The stems are straighter than in that species, and more or less woolly. The leaves are less strongly reticulate, quadrately oval, cordate at the base, distinctly and rather closely crenulate-serrate, woolly on each side but more especially above; petiole very short, and very woolly, as is the young shoot. The catkins are lateral ; the scales (not in good condition) seem to be larger ; cap- sule (burst) sessile, glabrous, shaped as in reticulata; style rather long; stigmas rather long, entire or bifid. This apparently very distinct plant I provisionally name S. soluta; but I am unwilling to place either it or the other in the list till they have been rediscovered. X SALIX SEMIRETICULATA, n. hybr. (S. reticulata x 8. ni- gricans ? ) A willow which was discovered at an altitude of 2300 feet on Meall Ghaordie, in Perthshire, by Mr. James Brebner, has a superficial resemblance to S. Grahami, chiefly from the habit of the plant and the shape of the leaves. It seems, however, to be a hybrid between S. reticulata and probably 5. nigri- cans, although it has not much resemblance to either of these species. It may be thus described :— Branches slender, long and trailing; bark fuscous brown glabrous and shining; shoots of the year dull paler brown white-pubescent. Buds red-brown, at first slightly pubescent REVISION OF THE BRITISH WILLOWS. 445 with white hairs, finally glabrous, quadrately-ovate and obtuse. Leaves at first half-folded, the older ones flatter, oblong-orbicular, truncate or subcordate at the base, obtuse or slightly twisted at the apex; margin slightly reflexed, slightly and rather remotely glandular serrate-crenate or nearly or quite entire, youngest leaves less obscurely crenate; surfaces at first, and especially below, more or less furnished with white woolly adpressed hairs, afterwards glabrous above and nearly or quite glabrous below; upper surface dark green, shining, reticulately rugose from all the veins, even in the smallest, being impressed (when dried, how- ever, some of the veins are raised); under surface dull paler green, white-dotted as is also the upper surface, the veins in- cluding the smaller ones somewhat thickened and reticulately prominent, more especially when young; veins numerous, pel- lucid; primary veins forming an acute angle with the midrib, those towards the base of the leaf more approximate; larger leaves about 1X4 inch, smaller 4X4 inch; petiole about À of the lamina in length, grey-woolly above; stipules (rare) minute, roundish, glandular, soon disappearing ; leaves involute in verna- tion. Catkins lateral on peduncles about as long as the catkin, and with three or four leaves with buds in the axils, the leaves similar to the other leaves but smaller, and the upper surface mostly glabrous, narrowly red at and near the apex; catkins ovate, short, compact; rachis stout, woolly ; scales membranous, brownish, clothed with long straight white hairs on both surfaces, broadly spathulate, obtusely rounded and erose at the apex; capsules (large for the size of the catkin) ovate subulate, slightly compressed obtuse (acute in the young ovary), green tinged more or less with red-brown, slightly hairy especially towards the top, but sometimes becoming almost or quite glabrous ; Pedicel silky or glabrous, stout, shorter than the long, linear, Very thin nectary ; style moderate, bifid at apex; stigmas about as long as the style, bifid, erect, but twisted and recurved at the apex *, * The above description was taken from the rather old catkins of the wild Specimens. The following is from a younger catkin from a cultivated plant:— les veined, broadly obovate, involvent ; apex truncate or slightly rounded, and sometimes slightly emarginate ; gieenish white at the base, brownish, reddish brown, or red in the middle (the upper scales being the brightest coloured), and black at the apex; upper part of both surfaces clothed with long white hairs, Ovary subulate-conic, shortly pedicellate, both ovary and pedicel with long hairs or somewhat glabrous below ; style stout, yellow, its L INN. JOURN.— BOTANY, VOL. XXVII. 2H 446 DR. F. BUCHANAN WHITE'S In some specimens the midrib and primary veins, as well as the petiole and margin, are somewhat reddish in colour. A willow which Mr. Brebner found on a neighbouring rock to the above much resembles it, but has flatter, more oblong leaves, more closely but shallowly crenate-serrate or almost serrate, more distinctly white-dotted above, less rugose above and less reticulate below, veins less pellucid, and margins less reflexed. I have not yet seen catkins *. In Smith's herbarium is a willow labelled in ink * Salix ellip- tica, nov. sp., Clova Mountains, Mr. Thos. Drummond, Mr. W. Hobertson, 1825," and in pencil *I have this as a rounder- leaved var. of S. Myrsinites.” It has no catkins, but—from memory—is much like S. semireticulata. I think that it has no connection with S. Myrsinites. X SALIX SIBYLLINA, n. hybr. (S. reticulata x S. Lapponum.) In Edinburgh University Herbarium are four small specimens of a willow found by Greville near Loch Brandy, Clova, m 1824. They are labelled “ possibly a pilose state of 5S. reticulata. Though without flowers, these specimens so evidently belong, I think, to a hybrid between S. reticulata and S. Lapponum, that I have ventured to give them a name and a place in this list- The twigs are intermediate (between the parents) in character, rather stout, shining, chestnut becoming grey; young shoots slender, somewhat pubescent. Leaves small (the largest 2X2 inch, but many are much smaller), ovate, ovate-oblong, or gub- rotund, at first woolly on each surface but more especially below, finally nearly glabrous above and slightly shining ; upper surface at first slightly impressed by reticulate veins ; under surface with primary veins raised, the others scarcely or slightly r aised, reticulate; margin incurved, with obscure glands as in reticulata ; apex obtuse, or with a short cartilaginous point as in Lapponum ; base subcordate, rounded, or shortly cuneate; petioles very short. Whilst bearing no striking resemblance to reticulata, it yet base hidden by the scale; stigmas equal in length to the style, thick, bifid, curved, and connivent, yellow. Nectary much longer than the pedicel of the ovary, linear oblong, rather thick, yellowish green. . * A cultivated plant has since this was written produced Q catkins which though not perfectly identical with those of the other plant, are not essentially different in structure. In the structure of the stigmas and scales there is much resemblance to S, Grahami. f : REVISION OF THE BRITISH WILLOWS. 447. shows in the venation, structure of the margin, and shape of the leaves, its relationship to that species. The specimens are too old to show the colour well, but that of the principal veins isin the direction of S. reticulata, namely reddish. From the look of the examples and the dead wood about them, they seem not to have been in robust condition, and probably the plant grows much larger. C. SYNANDRZ. Group 10. PURPURE X. 17. SALIX PURPUREA, L. Whilst both British and Continental salicologists distinguish several modifications or so-called varieties of S. purpurea, they are not altogether agreed as to what these are or as to the names they should bear. In Britain the “ species” into which Smith and Borrer divided 8. purpurea are still retained as varieties of that plant. These are Lambertiana, Sm., Woolgariana, Borr., and ramulosa, Borr. (sometimes included in Woolgariana); and according to Babing- ton, S. Helix, L. Andersson has, in addition to the type (a gracilis, G. and Gr. =8. purpurea, Sm.), Lambertiana, Sm. (including S. Woolgariana, Borr.), and S. Helix, “ L.” (which is not the S. Helix of British botanists, but seems to be the same as ramulosa, Bort.) Wimmer distinguishes the forms eriantha, gracilis, Lam- bertiana, styligera, sericea, and furcata. As characterizing their varieties, British botanists lay stress Upon the colour of the twigs and upon the form of the stigmas, Points which are almost or quite ignored by the Continental salicologists. That the colour of the bark is not 8 constant character may be learnt from an examination of almost auy living bush, when it will be seen that very frequently the colora- tion depends to a great degree upon exposure to the lisht. oreover, from authentically named specimens it would seem that even British botanists do not attach so much importance to this feature as the descriptions would imply. Thus in Smith's herbarium Lambertiana (from Lambert himself), whose twigs modern writers describe as purplish, has in two of the flowering- specimens dark-coloured, and in the third, pale, bark. 448 DR. F. BUCHANAN WHITES As for the shape of the stigmas, I think that not much reliance can be placed on that supposed characteristic. There remain therefore as primary characters, the form of the leaves, of the catkins, and of the capsules; and following Anders- son's views three chief forms may be distinguished by those who wish to retain varietal names. These are :— l. gracilis (or genuina), with slender catkins, small subovate capsules, and usually narrow leaves. 2. Lambertiana, with larger catkins, ovate-conie capsules, and larger leaves broad throughout (Lambéertiana), or con- spicuously broader above the middle (Woolgariana). 3. ramulosa (S. Helix, And., but since S. Helix, L., is dubious, the use of that name is not expedient), with rather stout capsules, and narrow, more elongate, and more acuminate leaves. After examining an extensive series, the conclusion I come to is that while extreme forms admit of having one or other of these “ varietal ” names applied to them, the various modifications pass insensibly (as Boswell-Syme remarks) into each other, and that there are many specimens of which it is impossible to say with cer- tainty to which variety they belong. Idoubt therefore the expe- diency of retaining any varietal names. Regarding Wimmer's forms (other than those mentioned above), there are amongst British specimens plants that might be referred to eriantha (catkins more hairy) though not in an extreme state, to styligera (ovary with a short style), and to sericea (young leaves more or less woolly). X SALIX RUBRA, Huds. (S. purpurea x S. viminalis.) Since there has never been perfect unanimity amongst salico- logists regarding Linné’s Salix Helix, I have used for the hybrid between S. purpurea and S. viminalis Hudson's common] y accepted name of S. rubra. At the same time it is by no means clear that S. Helix should not be the name to be adopted. Fries and Koch, and at a later period Andersson, thought that S. Helis was only a form of S. purpurea. Wimmer, on the other hand, believing that Linné could not have described under two names such a well-marked species (and one which he bad seen in a living state) as S. purpurea, makes S. Helix a synonym of his S. viminalis-purpurea. Smith and his followers considered S. Helis to be a distinct species; and the later British botanists (with the exception of Babington, who places it under purpurea) agree in thinking it to be a variety of rubra. REVISION OF THE BRITISH WILLOWS. 449 Now, whatever doubt attends the Linnean plant—and it is cer- tainly great,—there is none about Smith's. Both his descrip- tions and his specimens clearly indicate a hybrid between pur- purea and viminalis, though much more closely related to the former than to the latter. Itis to be noted, however, that many herbarium specimens named Salix Helix belong to S. purpurea. As a rule, S. rubra seems to show a smaller range of variation than most other hybrids, and hence its division by Grenier into two chief varieties—viminaloides and purpureoides—has some claim for consideration. But this supposed constancy of form is perhaps more apparent than real; and is due, in the first place, to the fact that in many cases the plants have been derived from cuttings, and not from seeds ; and, in the second, to the leaves of both parent species being of the linear-lanceolate type, and hence not affording so much scope for difference in outline in the hybrid. Var. viminaloides (or rubra genuina)is distinguished by its longer and narrower lanceolate leaves, more or less pubescent below, at least when young, and with the margins more or less revolute ; by the paler and more hairy 9 catkins; and by the longer style and stiginas. The petioles of the leaves are described as longer, but in this and other points there is variability. In some British books the filaments are said to be free except at the base; but in the same catkin filaments almost free or united to any point between the base and the apex may be found. Var. purpureoides (which includes Forbiana, Sm.) has leaves broader at the middle, with flat and more or less serrated margins and more entirely glabrous surfaces ; darker-coloured and less hairy catkins, and more obtuse capsules with shorter styles and thicker stigmas. The leaves are described as more shortly petioled than in viminaloides, and Continental examples often show this, but in authentic examples of the British Forbiana the reverse 18 the case, Andersson adopts Grenier’s division into two varieties ; but Wimmer, who distinguishes several forms, does not: and since ìn places where rubra occurs as a spontaneous hybrid, plants combining the characters of viminaloides and purpureoides may be found, I think that it is not desirable to maintain varieties in this hybrid more than in others. Wimmer’s forms are b. For- byana ; c. sericea, Koch (=S. eleagnifolia, Tausch), with the leaves clothed below with silvery-white hairs—a form which I ave seen in Perthshire; d. macrostigma ; and e. angustissima. These and other described modifications show that though, for 450 DR. F. BUCHANAN WHITE'S the reasons given above, there appears to be a constancy of form in Salix rubra, a series of conditions can be found connecting viminalis and purpurea, the chief grades being sericea, rubra, Forbiana, and Helix (of Smith). A Perthshire plant, otherwise intermediate in its characters, has rather remarkable stamens. Whilst the filaments are usually connate half-way or to near the top, in some of the flowers one or both are divided at the apex—the division being of unequal length, and presenting the appearance of a stamen with 2, 3, or 4 irregular branches. Itis to be noted that as these branches form acute and not obtuse angles with the main filament, this does not seem to be a true case of what Hayne has termed cladostemmy. x SALIX SORDIDA, Kern. (S. purpurea X Ñ. cinerea.) As the name S. Pontederana, Schleich., is, according to A. Kerner, dubious, having been applied to hybrids of S. purpurea with several of the Capree, it appears advisable to use Kerner 8 name of S. sordida for the hybrid with S. cinerea. Andersson, indeed, retains the name Pontederana ; but he makes it include the hybrids formed not only with S. cinerea, but with S. Caprea, S. grandifolia, and S. aurita, since he thinks that there is no sure method of separating them. As, however, he has not united these species, it is scarcely justifiable to unite the hybrids if it is at all possible to distinguish them; and he himself has kept them separate as varieties. Of S. sordida (whose synonyms are A. subpurpurea-cineret, Kern., and S. cinerea-purpurea, Wimm.), Andersson, following Wimmer, describes two modifications, viz.: 1. cinerascens (nearer S. cinerea), and 2. glaucescens (approaching S. purpurea, and probably S. Pontederana of Koch). It is to be noted, however, that in other respects Andersson's and Wimmer's descriptions do not tally in every partieular; and that, moreover, Wimmer's published specimens do not always altogether agree with his de- scriptions, as, for example, in the length ascribed to the pedicel of the capsule. In the Woody Island near Perth, where both S. purpurea and S. cinerea abound, S. sordida appears to be not uncommon. But while forms quite intermediate between the parent species occur; the majority of individuals are in character much nearer s. a- nerea than S. purpurea—so near, in fact, that in the absence of REVISION OF THE BRITISH WILLOWS. 451 flowers, some of them would with difficulty be diseriminated from cinerea otherwise than as slight modifications. As belonging to Salix sordida I put all plants which, however like cinerea they may be, have the filaments of the stamens more or less united to each other. Connate filaments may, according to Wimmer, be distinguished from cladostemmie ones by the branches forming an acute and not an obtuse angle; and one of the best characters of hybridization with S. purpurea is the pre- sence of connate stamens. In these Woody Island plants an extreme degree of connation israre. Frequently the filaments are united fora little way above the base only ; but in the same catkin free, slightly united, and more distinetly connate stamens may all be found—a phenomenon which may be seen in S. rubra and other hybrids of S. purpurea. Other points in which these cinerea-like plants are variable are the shape, colour, and pubescence of the leaves, stoutness and pubescence of the twigs, shape of the catkins, and colour of the anthers. In all these particulars they approach, or recede from, S. cinerea by almost imperceptible gradations. Probably, as Wimmer suggests, the cinerascens modifications of S. sordida have a somewhat different origin from those which are nearer 8. Purpurea, having been possibly produced by purpurea 9 x cinerea 3, or, what seems more likely, by the crossing of sordida with cinerea, The 9 of the cinerea-like forms is a much more difficult plant to distinguish than the d , since of course we have not the assist- ance afforded by the connate filaments. I have indeed found some plants whose catkins are so like those of cinerea, that it is only by the resemblance of their leaves and habit to some d Plants of sordida that they can be referred to that species. In addition to the Woody Island plants, I have seen specimens from a 2 bush found at Dalmarnock, on the Tay above Dunkeld, by Mr. C. M'Inotsh, and have found both sexes on the banks of the Tay below Perth. Another willow from the Woody Island I at present include under S. sordida (as var. rubella), though I am inclined to sus- pect that it may be a cross of S. rubra (with which it grows) and S. cinerea (i. e. S. purpurea x S. viminalis x S. cinerea). In its leaves and habit is is near S. purpurea, but has free stamens. e Very beautiful catkins are small or moderate in size; the Pnopened anthers orange-red, but when burst yellow from the 452 DR. F. BUCHANAN WHITE'S pollen, and finally, when empty, fuscous; leaves oblanceolate, dark green and shining above, very glaucous below, soon quite glabrous, but at first more or less pubescent with brownish hairs ; branches slender and straight, glabrous almost from the first. X SALIX DICHROA, Déll. (S. purpurea x S. aurita.) To this hybrid belongs the willow published by Mr. Leefe (Sal. Exs. iii. No. 59) as “Salix Pontederana ? Schl.," and found by him at Rothbury in Northumberland. S. dichroa is very similar to S. sordida ; but the cinerea element in the latter is replaced by aurita. Mr. Leefe, who describes his plant as *a small shrub with declining branches," has published d specimens only; and these agree so sufficiently well with examples of the hybrid issued by Wimmer and by Kerner, that there can be no doubt about their identity. Mr. Leefe found in the same place a willow which he thought “might be the female;" but the catkins are much deformed, being, in fact, her- maphrodite monstrosities. Of this I have seen one catkin only ; and from it not much can be learnt. The leaves, however, re- semble those of the ¢ plant, but are rather nearer in character to S. aurita. XSaLIx Dontana, Sm. (S. purpurea x S. repens.) Salix Doniana was founded by Smith (‘English Flora, iv. p- 213. 47) on a willow “sent from Scotland, as British, by the late Mr. George Don ;" and though it has continued to be re- tained in our floras, it is probable that the majority of the later botanists had come to the conclusion that its claims to be con- sidered British were very slight. For my own part I had, at the time that this paper was read, decided to omit it from the list of British Willows, because, amongst other reasons, it had not come directly from Don into Smith's hands, but from Borrer, who received it from George Anderson, and in the transmission from hand to hand some particulars of its occurrence might have been accidentally lost. But I have now the pleasure of restoring it to a place in the list, having, during the past summer, found un- doubtedly wild specimens on the banks of the river Tummel, near Pitlochry, in Perthshire. The plants I found were growing with S. purpurea, and in the neighbourhood of S. repens. They are distinctly intermediate in character between the parents. The following notes were taken emer ` mme e REVISION OF THE BRITISH WILLOWS. 453 from living specimens :--Bush about 18 inches high. Bark of the twigs shining brownish green, of the shoots redder and slightly downy. Some of the upper leaves subopposite, after the manner of Salix purpurea. Leaves dull, rather dark green with white hairs above; below dull pale green, with more copious long white hairs; margin cartilaginous, slightly incurved, very finely and remotely serrate, with reddish glands ; veins pellucid, the chief veins impressed on the upper surface and raised on the lower. In shape the leaves are similar to those of the S. pur- purea amongst which it was growing (i. e. similar to S. Doniana, Sal. Wob. t. 85). Catkins ( 9 ) rather old, lateral, on a peduncle having three or four small leaves. Scales ovate spathulate, rounded at the apex, upper half black, with long white hairs. Ovary ovate conic, blunt; style distinct, though short ; stigmas very short, thick, and rather broad, semi-bifid; ovary white pubescent, on a pedicel about as long as the linear pale yellow nectary. Though the 9 plant ouly has been found in Britain, both sexes are known in Central Europe, where S. Doniana is widely distributed. Andersson mentions four forms or varieties, VIZ. a. latifolia, (3. lingulata, y. linearis, ò. leiocarpa. Wimmer de- scribes six forms, but does not give them names. The filaments of the g flowers are connate at the base only, or for a third or half their length. DESCRIPTION OF THE PLATES. Puate IX. s regarding the number of E diagrammati . uu ` € statement of the different view between 1762 and 1886. British Willows held at various periods À k The dates given are those of the publication of the Works Wee representing the views of the period (see p- 334). The shad , columns indicate the number of wumbered “ species,” and are divide into three portions—dark, medium, and light. The dark shaded , ecies ; the medium shade portion shows the number of true sp light shade the number number now recognized as hybrids ; and the ` of those once supposed to be species, but now considered to be oie merely forms or to have no claim to be admitted as British. ^ ` last column is given the estimate for 1890 of the true species ( di shaded) and of the hybrids (medium shaded) which have been found m Britain. LINN. JOURN.— BOTANY, VOL. XXVII. 21 454 DR. F. BUCHANAN WHITE'S Piare X. A diagram to iliustrate the various ways in which compound hybrids (i. e. hybrids in whose parentage are included more than two species) may be formed. Puare XI. A diagram showing the relationship of both Tribes and Species as regards the British simple hybrid Willows. The larger circles represent the Tribes, the smaller circles the Species, and the lines connecting the latter indicate that Hybrids between these species have been found in Britain. The circle containing the Tribe Diandrs has been divided into four sections to show the altitudinal range of the species, since that limits to some degree the number of hybrids. Section I. is strietly lowland, the solitary species in it not ascending above 1000 feet above sea-level. Section II. includes species which, while most common in the low grounds, yet ascend into the alpine region of Section III. Section III. contains the species which rarely, if ever, descend below 1000 feet, and most frequently occur only at a much greater altitude. Section IV. contains two species which properly belong to Section IIL, but which occasionally, though very rarely, descend far below the 1000-feet line. The Tribes Pleiandre and Synandre inciude lowland species only. INDEX. The Tribes are printed in eu 311. CAPITALS, the groups in alics, the species and hybrids of Salix in Roman type, and unnamed hybrids are doubly indented. Capree, 377. voy? Dianpre, 377, Salix Fragiles, 362. Nitidule, 432. Nive@, +2}, Pentandre, 359. Phylicifolie, 395. PLEIANDRA, 347. Purpurea, 447, Repentes, 339. Salix acuminata, Sm., 413, 420, adscendens, Sm., 392. alba, L., 370. alba x decipiens, 355. alba x fragilis, 271. alba x pentandra, 362}. alba x triandra, 354. alupecuruides, Zuusch, 333. alpestris, And., 441. ambigua, Ehrh., 392. amygdalina, Z., 347, 548. angustifolia, Wulf., 391. aquatica, Sm., 381, 389. Arbuscula, L., 392, 410. Arbuscula, Sm., 991. — o Arbuscula x phylicitolia, 417. Arbuscula x herbacea, 439. Arbuscula x Lapponum, 430. Arbuscula x Myrsinites, 426. arenaria, Z., 426. atrocinerea, Brot., 980. aurita, L., 382, 384. aurita X Caprea, 387. aurita x cinerea, 383. aurita x herbacea, 411. aurita X Lapponum, 429. REVISION OF THE BRITISH WILLOWS. Salix aurita X Myrsinites, 434. aurita x nigricans, 409. aurita X phylicifolia, 402, 405. aurita X purpurea, 452. aurita X repens, 392. aurita X viminalis, 414. aurita-Lapponum, Wimm., 429. aurita-nigricans, Heidenr., 409. aurita-viminalis, Wimm., 414. Aurora, Lestad., 392. bicolor, Ehrh., 403. bicolor, Forbes, 403. borealis, Fr., 400. cerulea, Sm., 370. Calodendron. Wimm., 414, 420. campestris, Fr., 400. canariensis, C. Sm., 379. Caprea, L., 385. Caprea x aurita, 387. Caprea X cinerea, 386. Caprea X cinerea X phylicifolia, Caprea X coriacea, 410. Caprea x nigricans, 406. Caprea X phylicifolia, 402. Caprea X repens, 394. Caprea X viminalis, 414. Capree x viminalis, 413. Caprea-dasyclados, Wimm., 420. Caprea-nigricans, Wimm., 408. Caprea-repens, Lasch, 394. Caprea-viminalis, Wimm., 414. capreola, J. Kern., 387. carinata, Sm., 410. cinerea, L., 378. cinerea X aurita, 383. cinerea X Caprea, 386. simer icifoli | 402, x Gaprea X phylicifolia, cinerea X Lapponum, 430. einerea x phylicifolia, 402, 403. cmerea X nigricans, 408. cinerea X repens, 393. cinerea X purpurea, 450. cinerea x rubra, 451. wel X viminalis, 414, ils. c: nerea-Lapponum, Wimm., 430. i erea-limosa, Lestad., 430. vine purpurea, Kern., 450. ane repens, Wimm., 393. Algo Viminalis, Wimm., 414, conformis, Schleich. contorta, Crowe, 347, BAS. coriacea, Forbes, 409. Oriacea x Ca cotinifolia, ane 10. Croweana, Sn., 398. cuspidata, Schultz, 360. aphneola, Tausch, 426. | Salix daphnoides, Vill., 310. dasyclados, Wimm., 4 15, 420. decipiens, Hoffm., 348. decipiens x alba, 355. dichroa, Doll, 452. Dicksoniana, Sm., 399, 412. Doniana, Sm., 452. eleagnifolia, Tausch, 449. elliptica, Sm., 446. excelsior, Host, 374. ferruginea, G. And., 414, 419. ferruginea, Forbes, 414. finmarchica, Willd., 392. firma, Forbes, 407. foetida, Schleich., 410. Forbiana, Sm., 449. formosa, Willd., 411. fragilis, L., 363, 368. fragilis, var. britannica, B. White, 368. . fragilis, var. porcellanea, Baenitz, 5) 350. fragilis X alba, 371. fragilis x pentandra, 360. fragilis x triandra, Wimm., 348. fragilis-triandra, Wimm., 348. Friesiana, And., 391. fusca, L., 391. glauca, L., 426, 428. glauca, Sm., 426, 428. Grahami, Baker, 437, 444. grandifolia, Ser., 378. grisophylla, Forbes, 409. Hegetschweileri, Heer, 399. Lambertiana, Sm., 447 lanata, L., 421. lanata x herbacea, 494. lanata X reticulata, 422. Janceolata, Sm 200. l apponum, L., 426. L T pponun x Arbuscula, 490. Lapponum x aurita, 429. 456 Salix DR. F. BUCHANAN WIIITE'S Salix Lapponum x cinerea, 430. Lapponum x herbacea, 440. Lapponum x reticulata, 446. Lapponum-Arbuscula, Wimm., 430. Lapponum-herbacea, 441. latifolia, Forbes, 406. laurina, Sm., 402. laxiflora, Borr., 403. longifolia, Host, 420. ludificans, B. White, 402, 405. lutescens, 4. Kern., 383. Macnabiana, MacGillivray, 434. margarita, B. White, 441. marrubiifolia. Tausch, 426. Micheliana, Forbes, 418. mollissima, Ehrh., 355, 358. monspeliensis, Forbes, 369. montana, Forbes, 364, 376. Moorei, “Watson, L. C.," 438. multiformis, Doll, 355. myricoides, 404. Myrsinites, L., 432. Myrsinites x Arbuscula, 436. Myrsinites X aurita, 434. Myrsinites x nigricans. 433. Myrsinites x phylicifolia, 434. Myrsinites-nigricans, Wimm., 433. myrsinitoides, Fr., 433. nigricans, Sm., 396, 400. nigricans X aurita, 409. nigricans X Caprea, 406. nigricans X cinerea, 408, nigricans x herbacea, 438. nigricans x Myrsinites, 433. nigricans x phylicifolia, 401. nigricans x repens, 394. nigricans x reticulata, 444. nigricans-repens, Heidenr., 394. nigricans-Weigeliana, Wimm., A01. nivalis, Hook., 443. norvegica, And., 441. obscura, Doll, 419. oleifolia, Sm., 381, 384. palustris, Host, 374. pedicellata, Desf., 378. pendula, Ser., 364, 376. pentandra, L., 359, 365. pentandra x alba, 361. pentandra x fragilis, 360. pentandra x phylicifolia, 361. phylicifolia, L., 396. phylicifolia x Arbuscula, 412. phylicifolia x aurita, 402, 405. phylicifolia x Caprea, 409. phylicifolia x cinerea, 402, 403. phylicifolia x cinerea x Caprea, 402, 406. phylicifolia x herbacea, 437. phylicifolia x Myrsinites, 434. phylicifolia x nigricans, 401. phylicifolia x pentandra, 361. phylicifolia x purpurea, 399. phylicifolia x repens, 395. — az phylicifolia-nigricans, Wimm., 397, 398, 400. plicata, Fr., 392. Pontederana, Koch, 450. ` Pontederana, Schleich., 450. procumbens, Forbes, 433. prunifolia, Sm., 410. , pseudo-stipularis, Lond. Cat., 414, 416. puberula, Döll, 408. punctata, Wahlenb., 438. purpurea, L., 447. — purpurea x aurita, 452. purpurea X cinerea, 450. " purpurea x phylicifolia, 399. purpurea x repens, 452. purpurea x viminalis, 448. ramulosa, Borr., 447. ` Reichardti, A. Kern., 386. repens, L., 389. repens x aurita, 392. repens x Caprea, 31M. repens x cinerea, 393. repens X nigricans, 394. ` repens x phylicifolia, 399. repens X purpurea, 452. repens X viminalis, 391. aq repens-myrtilloides, Wimm., 392. reticulata, L., 443. reticulata x lanata, 429. reticulata X Lapponum, 446. reticulata X nigricans, 444. retusa, L., 438. rhætica, Kern., 399. rosmarinifolia, L., 390. rosmarinifolia, Sm., 391. rubra, Huds., 448. rubra x cinerea, 451. rugosa, Leefe, 414, 41%, 419. Russelliana, Sm., 363, 369. Sadleri, Boswell-Syme, 422. saxetana, B. White, 434. ` Schraderiana. Willd., 390. sejuncta, B. White, 444. semireticulata, B. White, HA sericans, Tausch, 414, 41%. serta, B. White, 436. . sibyllina, B. White, 446. silesiaca, Willd., 378, 38t i. simulatrix, B. White, 439. Smithiana, Willd., 413. sobrina, B. White, 440. soluta, B. White, 144. sordida, Kern., 450. . ei. sordida, var. rubella, B. W Ate, 451 spathulata, Willd., 392. speciosa, Host, 353. cecal REVISION OF THE BRITISH WILLOWS. 457 Salix sphacelata, Sm., 386. splendens, Bray, 370. spuria, Willd., 430. Stephania, B. White, 424. stipularis, Sm., 414, 415. strepida, Forbes, 408. Stuartiana, Sm., 426. subdola, B. White, 354. subpurpurea-cinerea, Kern., 450. superata, B. White, 423. tenuifolia, Sm., 402. tenuior, Borr., 409. ephrocarpa, Wimm., 402, 406. tetrapla, Walker, 400. thymelsoides, Schleich., 411. rani, eng., 855, 358. iri P y 855, 358 triandra X alba, 354. triandra x fragilis, 348. triandra x viminalis, 355. undulata, Ehrh., 355. LINN. JOURN.— BOTANY, VOL. XXVIT. Salix vacciniifolia, Walker, 411. vaudensis, Forbes, 408. velutina, Schrad., 414, 418. venulosa, Sm., 410. viminalis, L., 413. viminalis X purpurea, 448, viminalis x repens, 391. viminalis x Caprea, 413. viminalis x triandra, 355. viminalis-purpurea, Wimm., 448. viminalis-repens, Lasch, 391. viridis, Fr., 303, 367, 371. vitellina, Z., 371. Wahlenbergii, And., 433. Waldsteiniana, Willd., 411. Wardiana, B. White, 402, 403. Weigeliana, Willd., 400. Woolgariana, Borr., 447. Synanpre, 447. Triandre, 347. Viminales, 413. 402 DR. F. BUCHANAN WHITE'S Consequently the proper position of a number of specimens must remain doubtful. That Salix nigricans hybridizes with Caprea and cinerea is ad- mitted by all salicologists; but hybrids with aurita have been denied by some, though it seems certain that they exist. Since, then, nigricans crosses with these three Capree, it seems but reasonable to suppose that phylicifolia should form analogous hybrids. Salicologists, however, allow only one such compound namely with Caprea. There occur in Britain, however, some willows which seem to show more or less distinctly the hybrid- ization of phylicifolia with cinerea and aurita. X SALIX LAURINA, Sm. (S. phylicifolia x S. Caprea.) XSaLix Warprana (Leefe, MS.) n. hybr. (S. phylicifolia X S. cinerea.) XSALIX LUDIFICANS, n.hybr. (S. phylicifolia x S. aurita.) X SALIX TEPHROCARPA, Wimm. (S. phylicifolia x S. cinerea X S. Caprea.) x Salix laurina. Regarding the position and limits of S. Jaurina, there has been much disparity of opinion. Whilst Borrer and W. J. Hooker thought that it was very distinct, Walker-Arnott was inclined to believe that it connects S. nigricans and S. phylicifolia, and that S. tenuior, Borr., and S. tenuifolia, Sm., were synonyms and varieties of it. Leefe thought that it could not be satisfactorily distinguished from S. phylicifolia, of which species Babington con- siders it as a variety. Boswell-Syme gives it as intermediate, and J. D. Hooker as a hybrid between 8. phylicifolia and S. Caprea. Andersson and Wimmer, on the other hand, have no doubt that it is a hybrid between pAylicifolia and Caprea ; though Wimmer says that, on account of the uniformity of the specimens which he had seen from various parts of Continental Europe, it is 4 question whether it should not be considered to be a distinct species,and nota hybrid. This constancy of form, he thinks, is due, however, to the fact that probably all the specimens have been cultivated from an English stock. Ihave seen too few European examples to venture to endorse Wimmer's opinion; but the British plants which have been referred to S. laurina are by no means uniform in their characters; and to this is probably due the differences of opinion amongst British botanists. This REVISION OF THE BRITISH WILLOWS. 403 absence of uniformity, moreover, arises from the fact that two, if not three, different hybrids have been confounded under the name of Salix laurina, Sm. Adopting Andersson's view that S. laurina, Sm., is a hybrid between phylicifolia and Caprea, the figure in Eng. Bot. t. 1806, the description in the third edition of that work, and the specimens published by Wimmer (Coll. no. 90) and Reichenbach (no. 2417) may be cited as illustrating the characters of the species*. In most respects it very greatly resembles S. phylicifolia (with which, Andersson says, it has often been confounded), but shows its relationship to S. Caprea, not only in the structure of the 9 catkins (the g is unknown), but in the subarborescent growth, and in the size (and pubescence when it is present) of the leaves. From the resemblance of the leaves, S. laxiflora, Borr., has been referred to S. Jaurina (as, e. g., in the * Student's Flora,’ and, with some doubt, by Andersson); and it is possible that it may be a state much nearer S. phylicifolia, though from the structure of the flowers it is more probably a form only of that species, x Salix Wardiana. In addition to plants which agree with S. Jaurina, Sm., as in- dieated above, others have been published which, while evidently closely related to them, seem to show affinity with S. cinerea rather than with S. Caprea. Amongst these are several published by Leefe, including that " received from Mr. Borrer many years ago as the plant of Smith” (Fase. ii. No. 38), as well as the following :— Sal. Brit. N o. 43, found near Richmond, Yorkshire, by Ward, and labelled “ T should refer it to aquatica. Borrer.” Of this An- dersson said “Mihi S. Jaurine forma;” and Leefeand Ward thought this a better verdict than Borrer's. The same plant was published 1n Sal. Exs.iii. No. 60, as S. phylicifolia, *S. lauring, Sm., proxima," With the remark that, though desiring to call it after the dis- ` à Eng. Bot, t. 1806 is cited by Smith under S. bicolor, Ehrh. (of which he 8 Mou, ern a synonym); but S. bicolor, Ehrh., is only 8. phylicifolia. rina » Forbes, t. 38, though it has Smith's description of his bicolor (lau- ) appended, seems to be a different plant; and Wimmer thinks that it “Presents phylicifolia, though the style is rather short. 404 DR. F. BUCHANAN WHITES coverer, Mr. Ward, he (Leefe) feared that it could not be satis- factorily distinguished from Salix lawrina. Specimens seem also to have been afterwards distributed with the MS. name * S. Wardiana.” Sal. Ess. i. “No. 3. S. laurina, Sm.” “sent me under the name of S. myricoides." Sal. Ess. iii. “No. 62. S. phylicifolia.” * S. laurine, Sm., proxima." * This connects No. 60 with S. Jaurina, Sm., No. 38, and S. myricoides, No. 3.” * That these specimens have a similarity to each other may be gathered from Leefe’s remark on the last (Sal. Exs. iii. No. 62), and that they also bear resemblance to S. cinerea, Borrer's note on No. 43 suggestively indicates ; though how that acute sali- cologist could refer the catkins to that species is rather puzzling. Leefe's specimens are all much nearer to S. phylicifolia than to S. cinerea ; but in Perthshire I have found several willows which complete the transition from one species to the other; and it 18 probable that a search in the localities whence Leefe’s specimens came would reveal similar intermediate forms. The Perthshire specimens show much greater resemblance to cinerea than to phylicifolia ; but that they are hybrids of cinerea with one of the Phylicifolie is demonstrated by the evident style, while the general glabrosity of the twigs, and to some extent of the leaves, indicates that phylicifolia rather than nigricans is the parent. Compared with the true S. laurina, this hybrid may be distin- guished by its smaller catkins and smaller leaves, the general appearance and shape of which are suggestive of cinerea. The pubescence of the underside of the leaves shows a combination of the characters of the hairs of phylicifolia and cinerea without the soft woolliness of that of Caprea. The capsules, which are somewhat variable, show no very great points of distinction, though usually smaller and without the yeilowish whiteness frequently exhibited by the Caprea hybrid. For the hybrid between S. phylicifolia and S. cinerea I have adopted Leefe's MS. name, S. Wardiana, since it commemorates a botanist who was a diligent and sagacious student of British * Besides these, Leefe published another, Sal. Brit. Ers." No. 73, S. lat- rina, Sm.,” found at Richmond, Yorkshire, by Ward. Of this, Andersson says that it recedes somewhat from the true 8. laurina, but is nevertheless near it. If it is laurina at all, it is very near phylicifolia, and I would rather refer the specimens I have seen to that species, REVISION OF THE BRITISH WILLOWS. 405 Willows. The name Salix Wardiana is, of course, used in a wider sense than that in which it was employed by Leefe. The d has not been noticed, and though it probably occurs, it is very possibly difficult to recognize. x Salix ludificans. In addition to the Willows just noticed, I have seen specimens from a few bushes, which, whilst undoubtedly very close to „S. phylicifolia, cannot be satisfactorily referred to that species, on account of their short styles and other peculiarities, which seem to indicate that they are of hybrid origin. The phylicifolia element in them is so predominating that it is difficult to deter- mine the other parent, but after considerable study I think it must be S. aurita. Since S. aurita flowers a little later than the other Caprea, it might be imagined that hybrids of it with the Phy- licifolie would be at least as frequent as those of the other Capree; but this is either not the case, or they have been overlooked. l For this hybrid I propose the name S. ludificans. As already indicated, all the specimens I have seen are very much nearer Phylicifolia than aurita, though, judging from other hybrids, it is probable that forms more remote from phylicifolia occur. The specimens and some of their distinguishing characteristics are as follows :— A bush found by Mr. C. M'Intosh on the banks of the Tay above Dunkeld has the catkins ( 9) quite intermediate between Phylicifolia and aurita; but the leaves are very near those of the former species, exhibiting, however, in their shape a sug- gestion of aurita. The scanty pubescence, moreover (present on some of the leaves only), shows an affinity with aurita, since the hairs are finer and softer than in phylicifolia and slightly crisped. The other specimens I have seen were collected in Caithness by Mr. J. Grant and sent to me by Mr. A. Bennett. The examples are, unfortunately, not in the best condition and the Material is scanty. They represent three bushes, of which one is Sand two?. One has the catkins ( 9 ) rather different from the above-mentioned Perthshire plant, but, like it, intermediate tween itg supposed parents, and has the young leaves (I have aot seen mature ones) apparently also intermediate. Another * a striking resemblance to some forms of the hybrid between 406 . DR. F. BUCHANAN WHITE'S Salix aurita and S. nigricans, but with the nigricans element re- placed by phylicifolia, and, though a different looking plant from either of the above, shows its hybrid nature in all its parts. The g, though, like the others, near phylicifolia, yet shows, in the small catkins and structure of the leaves, a relationship with aurita; but if it had not been for the 9 examples it would probably have been referred to phylicifolia. X Salix tephrocarpa. In connection with the hybrids of cinerea and phylicifolia, mention must be made of the Willow described by Wimmer under the name S. tephrocarpa. Of this one bush only seems to be known, and, though it grows in the Berlin Botanic Garden, its origin is quite uncertain. Both Wichura and Wimmer tried in vain to ascertain the parentage of this plant, and the latter finally came to the con- clusion that it might be a hybrid between Jaurina and cinerea. A Willow found by Mr. C. M‘Intosh on the banks of the Tay, above Dunkeld, agrees pretty well with the description of S. tephrocarpa ; but I have been able to compare the leaves only, which I owe to the kindness of Herr Hennings, Director of the Berlin Botanie Garden. The Berlin and Dunkeld leaf-examples, though not identical, have a fair resemblance to each other, and both show an undoubted relationship with cinerea. In other respects the Dunkeld plant bas some affinity with Caprea and, to a slight degree, with phylicifulia. The catkins (2) are large and very handsome, the white capsules contrasting strongly with the conspicuously black scales. The style is very short but present; and in this and some other points the affinity with phylicifolia is shown. In the meantime I can find no place for this form but as 4 hybrid between cinerea, Caprea, and phylicifolia, and possibly arising from cinerea x phylicifolia crossing with Caprea. X Butt LATIFOLIA, Forbes. (S. nigricans x S. Caprea.) Wimmer and, on his authority, Andersson consider that 5. latifolia, Forbes, t. 118, is a hybrid between S. nigricans and S. Caprea. Leefe, on the other hand, has published (Sal. Ezs. ii REVISION OF THE BRITISH WILLOWS. 407 Nos. 52 & 53), under that name and with citation of the plate, a plant which he says is only a form of Salix nigricans. Comparing Leefe’s specimens with Forbes’s plateand description, I find that they do not agree, and that whilst Leefe’s latifolia is nothing more than what he thought it to be, i. e. nigricans, Forbes's Plant is evidently a hybrid form. From the affinity between S. Caprea aud S. cinerea, their hybrids with S. nigricans frequently so much resemble each other that it 18 not easy to separate them. Wimmer relies on the yellowish- white much thicker capsules, thicker, shorter and broader catkins, the much broader, oval-subrotund young leaves, and the larger, broader, and more hairy old leaves, as characters by which to distinguish the best form of latifolia from hybrids of cinerea with nigricans, but remarks that some specimens show a departure from these points of distinction. S. latifolia seems to have been found in a very few places in Lapland, Sweden, and Germany. The only undoubted specimens I have seen are all * from Perthshire, where three bushes—in two widely separated localities—have been found by Mr. C. M'Intosh and myself. Of these, one is quite intermediate between S. Caprea and S. ni- gricans ; another greatly resembles Sal. Wob. t. 118 and has more affinity with the cinerea-nigricans hybrid, but also seems to be, beyond doubt, latifolia; and the third, in its longer styles, inclines More to nigricans. Of 8. latifolia the only was known to Forbes and to Wim- mer, but Andersson describes the d. ` Wimmer refers S. firma, Forbes, t. 106, and S. cotinifolia, Sm., F orbes, t. 114, with some doubt, to S. latifolia ; but Andersson thinks that they represent nigricans only. Of S. firma I have Seen specimens from Kew Gardens which, though not quite identical with the plant figured by Forbes, are probably the same ; they possibly represent another form of latifolia, but I think are rather cinerea x nigricans. The plant usually called in Britain cotinifolia does not appear to be the same as Forbes's, and 8 only nigricans. * e . . . Since this was written I have found in Edinb. Univ. Herbarium a specimen eege “S.W. corner of Duddingston Loch,” which must be referred to S. lati- 408 DR. F. BUCHANAN WHITES X SALIX srREPIDA (Schleich.), Forbes. (S. nigricans x S. cinerea.) Andersson uses the name S. puberula, Do, for the hybrid formed by S. cinerea with S. nigricans, because he and Wimmer think that S. vaudensis (Schleich.), Forbes, is, though probably the same thing, somewhat doubtful. Wimmer is also of opinion that S. strepida (Schleich.), Forbes, t. 100, bears more resemblance to the same hybrid than to anything else, though Andersson refers it to S. nigricans. Of S. strepida Y have received specimens from Kew Gardens which are sufficiently like Forbes's figure, and which seem, with- out doubt, a hybrid between cinerea and nigricans. Consequently, since the name is earlier than either vaudensis or puberula, I have adopted it for this hybrid. S. firma, Forbes, t. 106 (which Wimmer thinks may be S. Caprea-nigricans), and S. vaudensis, Forbes, t. 117, are, judging from specimens cultivated at Kew, forms of the same hybrid. Like other compound willows, S. strepida is subject to con- siderable variation, increased in this case by the variability of its parents. In addition to this, the intimate alliance, on the one hand, of nigricans to phylicifolia, and, on the other, of Caprea and aurita to cinerea, augments the difficulty of satisfactorily placing every specimen, and makes it impossible to draft such 4 description as will serve to identify the hybrid in every case- At the same time S. strepida, in many of its forms, has a facies of its own which, when once learned, should not fail to guide the student to a correct discrimination of the species. Wimmer and, following him, Andersson describe three forms— a. puberula (Doll), B. vaudensis (Forbes), and y. nitida, Wimm. From the examination of a number of specimens, I am unable to see any sufficient reason for maintaining these varietal names. Inits best forms, strepida combines the characters of its parents, but not unfrequently it exhibits more relationship with one than with the other. In some of its states the leaves so much resemble those of nigricans, that it is only by the shorter style and stigmas that any connection with cinerea can be suspected. In others the leaves, catkins, and capsules are so like those of cinerea that the relation with nigricans (or perchance with phylicifolia) is shown only by the rather long styles. In Continental Europe S. strepida has been recorded from ? REVISION OF THE BRITISH WILLOWs. 409 very few places only, perhaps because it has not been carefully looked for. In Britain it may not unreasonably be expected to occur in those parts of the north where its parents grow together, as they often do, since in Perthshire it is of wide, though not of abundant, occurrence on the banks of the Tay and some of its tributary streams. I have found both sexes, but have seen more ? than d plants, probably because the former are more easily recognized. In the south-east of Scotland Mr. A. Brotherston has found plants which probably belong to Salix strepida, though the material I have seen is scarcely sufficient for absolute certainty. X SALIX CORIACEA (Schleich.), Forbes. (S. nigricans x S. aurita.) In his notes on S. cinerea-nigricans (S. strepida), Wimmer says that it is hardly to be doubted that hybrids between S. nigricans and S. aurita occur. As examples of such he cites S. conformis, Schleich., in Herb. Willdenow, and thinks that Forbes’s figure, t. 119, of S. grisophylla (Schleich.), Forbes, also represents such a hybrid. Andersson, after quoting Wimmer’s words, says that the S. conformis referred to is represented by three specimens, of Which two belong to S. aurita and the third to S. cinerea, and that the specimens of S. grisophylla, published by Schleicher, Which he has seen, are, as regards some, S. nigricans, and, as regards others, S. cinerea. Under the name S. aurita-nigricans, Heidenr., Heidenreich has distributed specimens of a Willow found near Tilsit in russia; but I am not aware that any description of it has been published. Here and there, on the banks of the Tay between Dunkeld and Logierait, Mr. C. M‘Intosh and I have found plants which "ppear to be certainly a hybrid between S. nigricans and S. aurita. As might be expected, these bear a very considerable resemblance to 8. strepida, and some of them, indeed, might be nearly as well referred to that hybrid, except for a certain appearance suggestive H aurita rather than cinerea. Others, however, show a distinct combination of the characters of aurita and nigricans, and agree tolerably well with Heidenreich’s examples. Though Wimmer refers it to S. nigricans, | think that Forbes’s 410 DR. F. BUCHANAN WHITE’S t. 112, Salix coriacea, represents this hybrid. From the plate alone no certain conclusion could be derived; but specimens received from Kew Gardens, and others published by Leefe (under 8. nigricans), can scarcely be satisfactorily placed otherwise than here, though they do not exhibit the best form of the hybrid. In the plate both g and 9 flowers are shown, but I have seen the only. The flowers in the specimens show a more marked re- lationship with aurita than they do in the plate. If I am right in referring Forbes's plant, S. coriacea (Schleich.), Forbes, must be the name of the hybrid; otherwise it will have to be called S. aurita-nigricans, Heidenr. Like its close ally S. strepida, S. coriacea shows considerable variation. From that species the form of the leaves, recalling aurita rather than cinerea, the usually smaller and narrower 9 catkins, the shape and smaller size of the capsules, and frequently the colour of the scales are characteristic points which will serve to distinguish coriacea. So far, however, as our specimens go, it is more likely to be passed over as a form of nigricans, from which, however, careful examination will show its distinctness. A plant found on the banks of the Tay above Dunkeld by Mr. C. M'Intosh must be noticed here, since it seems to include nigricans and aurita amongst its parents. From certain points, however, in which it resembles the willow (from the same locality) already referred to as S. tephrocarpa, Wimm., I am inclined to suspect that it also includes S. Caprea and is a hybrid of that species with S. coriacea. 10. But AnBuscura, L. There is a divergence of opinion amongst botanists as to the situation of S. Arbuscula. Many, including the British, think that it should be associated with S. Myrsinites, &c.; but Anders- son, on account of the great similarity of the dwarfer forms of S. phylicifolia and the larger sub-alpine states of S. Arbuscula, considers that its place is amongst the Phylicifolia. À number of varieties or modifications have been described, but various authors have treated these differently. The varieties of the ‘London Catalogue’ are :—a. carinata, Su. b. fetida, Schleich. (=prunifolia, Sm.), c. venulosa, Sm., and Eta PNEU E * REVISION OF THE BRITISH WILLOWS. 411 d. vaccinifolia, Walker—all of which were at one time supposed to be distinct species. Wimmer has:—a. Waldsteiniana, Willd., 3. formosa, Willd., and y. feetida, Schleich. (to which he refers venulosa and vaccinifolia). Andersson has also three chief modifications—a. erecta (with three leaf-forms which include Waldsteiniana, formosa, and pru- nifolia); B. humilis (with two leaf-forms which include fetida, venulosa, and vaccinifolia) ; and y. thymeleoides, Schleich. (about which Wimmer is doubtful, and which Andersson thinks may be a subpilose condition). Both Wimmer and Andersson place Salix carinata, Sm., as a synonym of S. Arbuscula. Tf any varieties are to be retained, Andersson’s arrangement seems to be the best. S. erecta is distinguished by its more up- right and taller growth, its larger leaves, and less leafy catkins; humilis by its smaller size, more creeping habit, smaller leaves, and by the catkins which, when young, are subglobose and buried in leaves. But whilst plants occur which agree well with the definition of one or other of these modifications, there are many which can- not be placed in one more than in the other. I therefore think that it is inexpedient to adopt any varietal names. As for the British forms, there seems to be little doubt but that Walker- Arnott is right in saying that they cannot be satisfactorily dis- tinguished. There are some discrepancies in the various descriptions of S. Arbuscula. By some authors the style is described as elongate or long; by others as very short or mediocre, which in the majority of cases it is, The capsule is sometimes said to be sessile, whereas it has a pedicel of varying length but always much shorter than the rather elongate nectary, a character which distinguishes it from S. phylicifolia, whose pedicel is always longer than the short nectary. The stipules are described as absent or very rare; but small stipules may often be seen on the young “hoots. The coma of the seeds is sometimes described as having à reddish tinge ; but in all our specimens it is white. l Some other points in which S. Arbuscula varies may be briefly noticed. The leaves, which exhibit a considerable range in shape and size, vary also in the pubescence and colour. The underside -$ sometimes glaucous, sometimes green or with white dots ; Sometimes distinctly pubescent when young, and at others glabrous. young branches and buds are often silky, though frequently 412 DR. F. BUCHANAN WHITE’S described as glabrous. The scales vary in colour, pubescence, length, and shape (roundish, obovate, and oval) ; the style from nearly none to almost one third the length of the ovary, and, with the stigmas, in being thick or thin ; and the capsule in the amount of pubescence and colour. With all these variations there is, however, usually no difficulty in easily recognizing the species. But Dicxsonrana, Sm. (S. phylicifolia x S. Arbus- cula.) Although Wimmer was inclined to think that S. humilis, Willd., might be a hybrid between S. phylicifolia and S. Arbuscula, he does not, in the *Salices Europes,’ go the length of describing it as such, and Andersson says that it is a modification only of S. phylicifolia. Hybrids between phylicifolia aud Arbuscula might be expected to occur; but if they do they must be very rare. As mentioned under S. phylicifolia, it seems quite within the bounds of possi- bility that S. Dicksoniana, Sm., is such a hybrid. Though origi- nally found in “Scotland” by Dickson, and in Breadalbane by Winch, only cultivated specimens appear to be now known, and the characters of these, as well as those given in Smith’s description, are in many ways intermediate between phylicifolia and Arbuscula. As already stated (p. 399), there is some doubt whether Smith's plant and Forbes's are the same ; but the proba- bility seems to be that they are only different conditions of one species, and I have therefore retained Smith's name of S. Dick- soniana. As regards the specimens I have seen, the leaves might belong to either S. phylicifolia or to S. Arbuscula, though, perhaps, on the whole nearer the latter species. From S. Arbuscula, also, has been derived the small stature of the plant (*a foot high," Smith; “18 inches to 2 feet," Forbes), the small catkins (9) appearing with the leaves and with leafy peduncles, the small capsules, which in colour and aspect recall those of Arbuscula, and the shortish styles; but from phylicifolia the thicker catkins, more longly-pedicelled capsules with stronger pubescence, and stouter styles and stigmas. The scales are somewhat intermediate in colour and structure. Both Smith (who places it, in Engl. Fl., next S. carinata) REVISION OF THE BRITISH WILLOWS. 413 aud Forbes (next Salir prunifolia) seem to have had some thoughts of the reiationship of Dicksoniana to Arbuscula forms. Neither Smith's plate (in which the catkins are much too young) nor Forbes's (in which the figured leaves resemble those of nigricans) can be called very good. It is to be hoped that the wild state of Salix Dicksoniana will be rediscovered. Group 7. VIMINALES. 11. SALIX viMINALIS, L. . Though, like other species, liable to modification, S. viminalis Is, in Western Europe at least, one of the less variable willows, and has no varieties worth y of distinct names, though one or two have been described. Amongst these are var. intricata, Leefe (distinguished chiefly by the cloven, reflexed, and entangled stigmas), and var. stipularis, Leefe (with broader leaves and longer stipules). The 9 catkins vary a good deal in size. Most of the British and many of the foreign specimens I have seen have much shorter, but proportionately broader, catkins than in Wimmer’s published examples, but apparently do not belong to Déll’s var. abbreviata, in which the catkins are described as linear-oblong. XSatrx Surruraxa, Willd. (S. viminalis x the Capree.) The hybrids which S. viminalis makes with the Capree form à group the treatment of which is very difficult—a difficulty not diminisheg by the manner in which the group has been dealt with by botanists. ) For not only have different names been given to one and the “ame plant, but different plants have had the same name ascribed ‘ot em. Thus four, more or Jess distinct, forms have been called fcuminata,” and the true S. acuminata has been described under two other names. Much of this confusion has doubtless Teen, not Only from the great variability of the hybrids, but from the fact that the plants described by British authors were " familiar to the Continental salicologists, nor those of the atter to British botanists. No the 8roup is so specially a British one, it will be of interest later ne Its constituents according to the views of some of the T British botanists. It includes, according to :— LINN, JOURN.— BOTANY, VOL. XXVII. 2r 414 DR. F. BUCHANAN WHITE'S 1. W. J. Hooker and Borrer: S. stipularis, Sm., S. Smithiana, Willd., S. ferruginea, “ And., MS.,” S. acuminata, Sm., 8. holo- sericea, “ Willd.” ( 2S. acuminata, var. rugosa, Sm.). 2. Walker-Arnott: S. stipularis, Sm., S. Smithiana, WO, S. acuminata, Sm., S. ferruginea, And., S. holosericea, “Willd. ? 3. Leefe (Journ. Bot. 1871): S. stipularis, Sm., S. Smithiana, Willd., S. acuminata, Sm. 4. Boswell-Syme: S. stipularis, Sm., S. Smithiana, Willd. (with var. stipularis), S. ferruginea, And. (with var. rugosa), S. acum- nata, Sm. 5. Babington (1881) : S. stipularis, Sm., S. Smithiana, Willd. (with varr. rugosa and ferruginea), S. acuminata, Sm. 6. J. D. Hooker (1884): As supposed hybrids of S. viminalis— S. stipularis, Sm., S. Smithiana, Willd., S. acuminata, Sm., S. ferruginea, G. And., S. holosericea, Willd., S. rugosa, Leefe (7 holosericea, Hook. & Arn.). 7. London Catalogue (1886): S. stipularis, Sm., S. Smithiana, Willd. (with var. pseudo-stipularis, Lond. Cat.) S. ferruginea, G. And., S. rugosa, Leefe, S. acuminata, Sm. Wimmer has as the constituents :— S. Caprea-viminalis, Wimm. (=S. Smithiana, Willd.). S. cinerea-viminalis, Wimm. S. aurita-viminalis, Wimm. S. stipularis, Sm. S. Calodendron, Wimm. (—S. acuminata, Sm.). S. holosericea, Willd. Andersson praises Wimmer's arrangement, but remarks that authors have described individual specimens rather than the main forms themselves. His own arrangement is as follows. I have added within brackets the equivalents in Wimmer’s * Salices’ and in the * London Catalogue ' :— S. stipularis, Sm. (also of Wimmer and Lond. Cat.). S. Smithiana, Willd. «. sericans (Tausch). (S. Caprea-viminalis, Wimm. S. cinerea-viminalis, Wimm., p. P- S. aurita-viminalis, Wimm., p. P: S. Smithiana, Willd., Lond. Cat-) B. velutina (Schrad.). (S. cinerea-viminalis, Wimm., p- P S. holosericea, Willd., Wimm. S. rugosa, Leefe, at least p. p) REVISION OF THE BRITISH WILLOWS. 415 Var. ferruginea (Forbes). (S. cinerea-viminalis, Wimm. S. ferruginea, G. And., Lond. Cat.) y. acuminata (Sm.). (S. Calodendron, Wimm. S. dasyclados, Wimm. S. acuminata, Sm., Lond. Cat.) The weak point of this arrangement is that under one specific name (Sali Smithiana) several hybrid plants—having one only of the parents in common—are included. If S. Caprea, S. einerea, and 8. aurita are retained as distinct species—and Andersson did not unite them,—their several combinations with S. viminalis should necessarily, according to the method adopted by Andersson in the case of other hybrids, have separate recognition. On the other hand, comes the almost unsurmountable difficulty of distin- guishing between them, especially when, as in several cases, the exact parentage has not been proved, and when it is by no means improbable that a second hybridization, either with one or other of the parents or with a third species, has taken place. For this reason I am inclined to adopt provisionally the following modifi- cation of Andersson's arrangement :— S. Smithiana, Willd. a. stipularis (Sm.). B. sericans (Tausch). Y- velutina (Schrad.). d. Serruginea (G. And.). €. acuminata (Sm.). dat re proceeding to notice each of these varieties, I may say cimene ( e result of the study of a very large series of spe- tinental TH of them authentically named), both British and Con- will Wë , are failed to find such a permanence of characters as examples ° definitely separate one form from another. Certain under on can be, without much hesitation, placed satisfactorily ever that or other of the varietal names. There are many, how- anoth o cannot really be referred to one variety more than to of each , bi which, combining to some extent the characteristics said tha orm connecting-links. Of some others little more can be an that they are modifications of S. Smithiana. ie a. stipularis. (Salix stipularis, Sm.) d . ` pe the statement given above, it will be seen that most of the cited retain S. stipularis as a species. Even Andersson, Who . has united the other forms, gives it separate rank. But 2r2. 416 DR. F. BUCHANAN WHITE'S from a study, both of descriptions and specimens, I cannot find any grounds for giving Salix stipularis a higher position than that which some of the other forms (as, e. g., S. acuminata) have been considered entitled to. Of all the varieties, stipularis is the one which is most closely related to S. viminalis. From that species it may be distinguished by the broader leaves with less shiniug pubescence on the under surface, longer and broader stipules, larger catkins, and less sessile capsules; but whilst these are its normal characteristics, they may be all more or less so modified as to afford gradations into S. viminalis on the one side or into (9. sericans on the other. From the latter, its usually tomentose twigs, larger and darker catkins, more sessile capsules, and longer and more filiform stigmas serve to distinguish it, as do also, in the more typical examples, the shape and size of the leaves and of the stipules. But the latter are subject to modification and are not always present; and the former, which vary even in the same specimen, are not unfrequently so similar to those of some conditions of sericans that, were it not for the other characters, some examples would certainly be, and frequently are, called sericans. In a word, there is no sharp boundary between S. viminalis and stipularis nor between stipularis and sericans. Willows more or less resembling stipularis have been named S. viminalis var. stipularis, Leefe, and S. Smithiana var. pseudo- stipularis, Lond. Cat. Of these the first includes mere modifi- cations of S. viminalis, as well as forms of stipularis approaching viminalis; whilst “ pseudo-stipularis" seems to be applied to plants connecting stipularis with sericans. The d of stipularis seems now to be unknown * though it is figured both by Smith and by Forbes. The parentage, other than from S. viminalis, is doubtful, and some authors have thought that S. cinerea is the unknown factor in its origiD- . Wimmer suggests that it is a hybrid between S. viminalis and S. dasyclados, Wimm. (=S. acuminata, Sm.). I suspect that Wimmer is so far right in supposing that it has originated from a cross between S. viminalis and one of the viminalis-Capreé hybrids; but whether this hybrid is acuminata is doubtful. ` S. stipularis is almost confined to Britain, but, though 1 value as an osier is said to be small, whether it is really more than an escape from cultivation is very uncertain. * Sincethis was written I have found a single g bush—apparently self-sown— near Perth. I fear, however, that it has been lately eradicated. REVISTON OF THE BRITISH WILLOWS. 417 B. sericans. (Salix sericans, Tausch.) This, which includes “ Salix Smithiana as generally understood, is a most variable plant, and is almost inseparably connected with stipularis on the one side and velutina on the other. From the former, its smaller and paler-coloured catkins, longer- stalked capsules, and usually shorter and thicker stigmas, and from the latter its rather larger catkins and shorter-stalked capsules, are the best points of distinction ; whilst another, which separates it from both, is to be found in the usually glabrous twigs. But in any large series of specimens all these points will be found to be subject to much variation. The leaves are very variable, not only in shape and size, but in the nature and amount of the pubescence. In shape the leaves pass—even in the same example—from ovate-oblong to linear- lanceolate; whilst the pubescence of the underside may either be silky or woolly, shining or dull, copious or almost altogether absent. Asa rule, the margin is entire or very slightly crenate ; but in forms which approach velutina it is more evidently crenate or even subserrate. These latter forms Andersson places in modification 3. subobscura (of sericans). Some of the forms, as 1. latifolia, Aud., and 2. angustifolia (Wimm.), And., are, as Wichura has proved by experiment, hybrids of S. viminalis with S. Caprea; but the origin of others (as, e. g., 3. subobscura, And.) is not so clear. Under subobscura Andersson cites, as belonging in part, Wim- mer’s S. cinerea-viminalis and S. aurita-viminalis ; and from the characteristics of this form it seems very probable that S. cinerea and S. aurita are concerned in the production of some of its Modifications. Some of the specimens which have been referred to S. rugosa, Leefe, should probably find a position here. To these forms Andersson adds a fourth—serotina, which, however, he thinks is only accidental. It is distinguished by the catkins being shortly but distinctly stalked, and the peduncle furnished with small leaves. Many of our 9 plants have catkins of this nature. Var. sericans seems to be the most common of the varieties of 8. Smithiana, but appears to occur more frequently as a cul- tivated plant or as an escape from cultivation than as a truly wild Ze, The most abundant form of it is 2. angustifolia, or near it. It presents, however, great variability in small particulars. 418 DR. F. BUCHANAN WHITE'S y. velutina. (Salix velutina, Schrad., S. holosericea, Willd., S. cinerea-viminalis, Wimm. p. p., S. Micheliana, Forbes, Sal. Wob.) Var. velutina, which is often closely connected with sericans, and especially with the form subobscura (from some states of which it can scarcely be separated), is distinguished by its usually tomentose twigs, serrulate leaves, shorter style and stigmas, and more longly pedicelled capsule. Whilst in stipularis the capsule is scarcely stalked, and in sericans the stalk is about as long as the nectary, or at most twice as long, in velutina it is twice OT three times the length. Just as in the other varieties, however, all the characters are liable to modification; so that it becomes impossible to say where certain specimens should be placed. Though the leaves are normally more or less serrate, they are not invariably so. Forms of this group which have serrate leaves seem to have derived their origin from the hybridization of S. cinerea or S. aurita, rather than S. Caprea, with S. viminalis. This appears to be in a measure indicated by the position of the marginal glands. In S. viminalis the position of these is some what remote from the entire margin. In S. Caprea they are also rather remote from the margin, which, however, is by n9 means always entire. In S. cinerea and S. aurita the glands are much nearer, if not on the margin, and are often at the apex of teeth. The position, therefore, of the glands and nature of the margin seems to afford a frequent (though not a constant) guide to the possible origin and position of some of the varieties of S. Smithiana, such as velutina and Serruginea. Var. velutina, which, in a greater degree than all the other varieties, appears to have arisen from the crossing of S. viminalis with S. cinerea, is said to be everywhere rare. Specimens which scem to be referable to it I have seen from Kelvington, N. York- shire (J. G. Baker), named S. cinerea-viminalis, Wimm., and well marked. Less well-marked are examples from Llangarren, Herefordshire (A. Ley); and Pendeford Mill, Staffordshire, named S. rugosa (Dr. Fraser). Dr. Fraser has also found. plants near Wolverhampton, which both Leefe and Boswell-Syme have referred to S. holosericea, Willd., Koch (=S. velutina, Schrad. sec. And.). These, which are g, are certainly rightly placed under or near velutina, but they are not very like Wimmer’s examples of S. holosericea, Willd. ; and the flowering-twigs and buds are more glabrous than Koch's comparison of them with those of S. cinerea REVISION OF THE BRITISH WILLOWS. 419 would imply. Koch says that the catkins, buds, and twigs of holosericea bear so great a resemblance to those of cinerea, that it. is impossible to separate the two plants except by the leaves. Some of the plants named Salix rugosa, Leefe, seem to properly belong to velutina. ^ Leefe's own specimens (Sal. Brit. Ess. No. 30, and Sal. Exs. i. No. 22)—of which he writes, “I find it difficult to say whether this plant is nearer to S. holosericea, W., or S. Smithiana, W.; with much hesitation I have thought it on the whole to belong to the latter"—are very like Wimmer's example (Coll. Sal. No. 100) of holosericea, Willd., except that Wimmer's is more pubescent. (Andersson remarked of the Sal. Brit. Ers. No. 80 specimens that they seemed to him to be S. acuminata, Sm.; but Leefe justly says that they do not at all resemble S. acuminata, Sm. Eng. Bot., and Ward declares them to be true “ holosericea, Hook.") Between Dr. Fraser's holosericea and the plants from Pendeford Mill, named by him rugosa, I see no essential difference in the leaves. The latter are 9 plants, and from the tomentose twigs &c. seem to be velutina. So also is & “rugosa” from Clevedon (J. W. White); but other plants, from various sources, named rugosa pass by degrees into sericans f. subobscura. 8. ferruginea (Salix ferruginea, G. And.). This, which Andersson makes a variety of his y. velutina, resembles an intermixture of S. aurita and S. viminalis, though Whether it is really a hybrid of these species requires proof. Wimmer quotes English specimens (from Piuley, T. Kirk), named S. ferruginea, under his S. cinerea-viminalis ; and the examples I have seen from the same locality and collector seem probably referable to velutina, which, as has been already mentioned, Includes in part S. cinerea-viminalis, Wimm. Var. Jerruginea is normally a smaller plant in every respect than "LY of the foregoing varieties, and has more slender and more glabrous twigs, smaller leaves, usually broader above the middle and almost destitute of pubescence, and smaller catkins, An- ersson points out that ferruginea may be the same as S. obscura, Dill, a plant very difficult to distinguish from the smaller states of the modification subobscura of sericans, from which statement t may be gathered that that form is one of the links between erruginea and sericans. Boswell-Syme makes S. rugosa, Leefe, a Variety of Serruginea, distinguished by the different style and 420 - DR. F. BUCHANAN WHITE’S stigmas and more pubescent leaves—all very variable characters 1n this group. Salix rugosa, as said already, I prefer to place under velutina, though often passing into sericans. Some states of plants which I believe to be ferruginea very much resemble Wimmer's examples of S. holosericea, Willd. ; and just as sericans passes into velutina, so does the latter, I think, into ferruginea. Herbarium specimens named ferruginea are rather a mixed lot, including not only S. cinerea and velutina, but even perhaps sericans. Leefe’s ferruginea (Sal. Exs. iv. No. 89), received from Woburn, seems to be the same as N. J. Andersson's ferruginea; and Leefe says that it is G. Anderson’s plant, but not that of Eng. Bot. On the other hand, another ferruginea published by Leefe (Sal. Brit. Exs. No. 35, and Sal. Ess. iii. No. 63), brought from Essex, is rather a different plant. Leefe says that it is the same as Eng. Bot. Sup. t. 2665, but I cannot say that I see the resemblance. Andersson said of it that it was nearest to S. holosericea, Willd., and it seems to be very close to if not identical with velutina. . Many of the plants called ferruginea by English botanists have the leaves densely hairy below, and, apart from the other differences which they display, cannot well be the same as the plant defined by Borrer as minutely hairy. In a living and wild condition, ferruginea is known to me in one locality (in Perthshire) only. Here it has all the appearance of being a natural hybrid between S. viminalis and S. cinerea, since S. aurita does not occur in the immediate neigh- bourhood. I suspect that here, therefore, it is really a hybrid between viminalis and cinerea, with a greater proportion of the latter than of the former in it; aud that, since the common British form of S. cinerea is unlike the common Continental form, ferruginea has not been so readily recognized as a state 0 S. cinerea x S. viminalis. e. acuminata (Salix acuminata, Sm.). Wimmer, thinking that the name acuminata, having been variously applied, is ambiguous, replaced it by his own name Calodendron, which, on account of the doubtful origin of the plant, he prefers to his earlier Caprea-dasyclados. Andersson quotes S. Calodendron, Wimm., as a synonym of A acuminata, Sm., and says that the latter is elosely allied to, if not identical with, S. dasyclado, Wimm. (=s. longifolia, Host sec. Wimm.). REVISION OF THE BRITISH WILLOWS. 421 Comparing Leefe’s specimens (Sal. Brit. Exs. No. 37, and Sal. Exs. ii. No. 27) of Salix acuminata with Wimmer's dasyclados (Coll. No. 99) I find that they are practically identical. Wim- mer’s exponent (Coll. No. 100) of acuminata is also like Leefe's specimens, but with a somewhat greater look of Caprea. Of Leefe's No. 37, Andersson remarked that it was most certainly dasyclados, and very markedly distinct from acuminata, Sm. ; but Leefe maintained that it was the acuminata, Sm., Eug. Bot., and cites Borrer as confirming the name. From a capsule preserved with Smith's original drawings, I have no doubt that Leefe is right as regards his plant, and I have also no doubt that dasyclados is a synonym of acuminata. From the specimens which I have seen I am rather inclined to think that acuminata deserves a more distinct position than the other varieties of Smithiana; but as Andersson has placed it among them, I do not venture to remove it. Its large catkins, densely hairy with crisped hairs, and the long erect woolly pubescence of the capsules, make it easily recognizable. The stout twigs are also most generally densely tomentose, and the large leaves are usually broader upwards and glaucous below. Andersson defines three modifications—glabrescens (to which most of the British specimens belong), virescens, and cinerascens (also British). The parentage of acuminata is doubtful. Andersson, by saying that its characters are intermediate between S. viminalis and S. Caprea, seems to think that it has sprung from these species. But Wichura has proved by experiment that sericans is produced by the union of viminalis and Caprea ; and it is difficult to believe that in acuminata we have not a different combination, arising perhaps from a second hybridization with S. Caprea. From the dubiety which involves the parentage of most of the varieties of Smithiana, a series of experimenta in hybridizing S. viminalis with S. Caprea, S. cinerea, and S. aurita—both sepa- rately and in combination with the resulting hybrids—is much to be desired. Till this is done it seems hopeless to expect that the various forms can be extricated from the confusion in which they are at | resent. Group 8. NIVES. 12. SALIX LANATA, L. This handsom2 species see.ns to be a little more variable than 422 DR. F. BUCHANAN WHITE'S the British descriptions of the plant generally indieate. The leaves vary a good deal in size; in shape from suborbicular to oblong-obovate ; and in the amount of pubescence from being per- sistently woolly to nearly quite glabrous almost from the beginning. The catkins vary in size and in the eclour of the pubescence of the scales, which, though usually yellow, sometimes fading to grey, is occasionally grey from the first. ' Andersson describes several modifications, but there are none which deserve varietal rank. Salix lanata seems almost confined (in Britain) to Forfarshire and Aberdeenshire, or at any rate is most common in these counties. Whilst recorded long ago from Perthshire, it seems to have been lost sight of there till recent years, when it was redis- covered by Messrs. Brebner and Haggart and myself in a few places in the Glen Lochay hills. Var. Sadleri (Boswell-Syme). The plant described by Boswell-Syme as Salix Sadleri has now been in cultivation for several years, and has developed some features not seen in the original wild specimens. The stems have increased in thickness and become tolerably stout; the leavee have become larger (some being1}x1#inch). M any of them are more decidedly cordate at the base, and many are also furnished with large ovate-acute glandular-toothed stipules. Though the margins of the leaves were described as entire, yet even in the original specimens the edges of some of the leaves are, more especially towards the base, very finely glandular-serrate or crenate, the teeth, however, being almost reduced to glands. In mature leaves the margin is recurved or thickened. The catkins, which, so far as I have seen, have not increased much in size, may be described as either terminal or as terminal and lateral, according to the view taken of the structure of their peduncles. They are situated at the end of shoots which have two or three leaves. These shoots might be considered to be leafy peduncles, but as their leaves are furnished with stipules and have buds in their axils they are really perhaps a permanent part of the plaut, and in this case the peduncle of the catkin must be described as leafless. The scales are oblong and concave, with the tip rounded, emarginate or truncate ; in colour they are greenish, with the apex occasionally tinged with red or very shortly black-tipped, and are clothed with long white hairs, The greenish-yellow REVISION OF THE BRITISH WILLOWS. 423 style is about as long as the naked young ovary; the yellow stigmas are bifid and spreading, and half as long as the style ; the nectary and the pubescent pedicel of the ovary resemble those parts in Salix lanata. S. Sadleri has been supposed to be a hybrid of S. lanata and S. reticulata ; but after long and repeated study of both wild and eultivated specimens, I cannot recognize in it any trace of S. reticulata, nor of any species but S. lanata, a remarkable form of which I think it must be considered to be. Serrate leaves are not common in 5S. lanata, though Andersson mentions their occurrence, and I have seen examples. Moreover, in S. Sadleri the serration is obscure and not constant. The chief differences lie in the small catkins and in some parts of their struc- ture, but even in these the Janata element is preeminent. The whole facies of the plant is that of a small form of S. lanata. X SALIX suPERATA, n. hybr. (S. lanata x S. reticulata.) A willow which grows, in company with S. lanata and other mountain-species, on the rocks at the head of Allt Innis Choarach, Glen Lochay, Perthshire, has required a considerable amount of study to decipher. At first sight the facies of it does not distinctly suggest any particular affinities ; but on examination, and keeping im mind the species with which it is associated, a clear relationship to S. lanata is revealed. The other parentage seemed very difficult to guess, but a certain uudescribable appearance of the plant hinted that it might be found in S. reticulata. The possibility of this suspicion being correct further study has made a probability. The following description is taken from dried specimens :— A very low bush, with erect or ascending branches ; stems rather stout, twigs moderate. Bark rich brown, rather shining ; young shoots greenish, sparingly hairy, but soon glabrous. Buds shortly oval-oblong, obtuse, at first greenish and very slightly hairy, then pale brown and glabrous. Leaves more or less obovate, base attenuate and often unequal, apex shortly pointed and plicate- twisted ; leaves at the top of the shoots the largest (average size about 14 x 2 of an inch), those below smaller and proportionately narrower; upper surface yellowish green, scarcely shining, flat, with slightly impressed veins; lower subglaucous dull, with the chief veins conspicuously raised and reticulate; upper surface slightly woolly with short hairs, under with long and short hairs, 424. DR. F. BUCHANAN WHITE'S both soon nearly glabrous; margin slightly cartilaginously thick- ened, entire or with a few minute glandular teeth, especially towards the base; petiole medium in length, channelled above, more or less tinged with purple, which colour sometimes also spreads to the midrib, veins, and margin, though these are more frequently yellowish ; stipules, when present, rather small, ovate, glandular-toothed. Catkins (d) terminal on the branches, sessile, short but stout (4 x inch) ; scales roundish, spathulate, rather broad, thin, brown with darker veins, densely pubescent on both surfaces with long straight white hairs; filaments glabrous; nectary half surroundiug the base of the filaments, short, split into 2 or 3 obcuneate pieces. From Salix lanata the more erect growth, shape of the leaves, and sessile woolly catkins seem to have been derived; whilst to S. reticulata the smaller size and glabrosity of the leaves, the brighter green of the upper surface, the colour of the petioles, veins, and margins, the rounded scales of the catkins, and above all the structure of the nectary, appear to be due. The elevated reticulation of the underside of the leaves is not greater than that which may often be seen in S. Janata, but the faint glandular toothing is very similar to what is shown by some specimens of S. reticulata. I have seen other plants which may alao prove to be hybrids of S. lanata and S. reticulata, but judgment on them must be reserved till a larger series of specimens has been obtained. X Sarx STEPHANIA, n. hybr. (S. lanata x S. herbacea.) Near Coire Dhubh Ghalair, Glen Lochay, Perthshire, Mr. D. A. Haggart found, two or three years ago, a very small willow whose affinities at first sight were rather difficult to determine, but which after much consideration I concluded to be with S. lanata and S. herbacea, in whose company it grows. More recently I have found in the same locality another form of the plant, and its cha- racteristics seem to point pretty clearly to a parentage from these species, It may be expedient to describe both forms. Though I have seen the original plant growing, the following description of it 18 taken mostly from dried specimens. Stems more or less buried ; branches short, moderately slender, and rather tortuous ; young shoots downy, but soon becoming -——— REVISION OF THE BRITISH WILLOWS. 425 glabrous. Leaves nearly orbicular, slightly cordate at the base, slightly longer than broad (the largest about Z inch long); margins more or less crenate-serrate; surfaces bright pale green, but not shining, slightly hairy with white woolly pubescence, more especially on the margin; veins pellucid, anastomosing, above impressed (when young) or slightly raised (when old), below ele- vated ; stipules minute, gland-like. Catkins (large for the size of the plant, being 3 inch in length) rather lax-flowered, on lateral peduncles which are furnished with one or two leaves; scales small, roundish, fuscous-brown but paler at the base, clothed with long white hairs; ovary lanceolate-conic, rather obtuse, glabrous except at the extreme base and on the very short pedicel (thus resembling some forms of S. herbacea); pedicel equal to or shorter than the nectary ; style moderately long, slender, purple (whieh colour occasionally tinges the ovary also); stigmas slender, bipartite. The more recently found form was described when living. It makes a low plant, with flexuous slender rooting stems buried amongst moss. Bark of the twigs rich but pale brown (becoming dark when dried) or greenish brown, glabrous or slightly downy ; young shoots greener and sparsely pubescent with long white hairs; buds rather long, oblong, blunt, at first green and hairy, then brown and glabrous. Leaves wavy and balf-folded, from orbieular to oblong in shape (the largest 14x14 inches); base cordate, rounded, or slightly attenuated; tip often shortly pointed and twisted; upper surface rather dark green and somewhat shining, lower paler and subopaque, both with numerous minute white dots, which disappear more or less in the dried plant; youngest leaves more or less clothed with long white hairs, but soon becoming glabrous; veins pellucid, reticulate, impressed above (or slightly raised in the older leaves when dried) and elevated below ; margin crenate-serrate, pinkish (as are the veins) in the youngest leaves (this colour disappears in the dried plant) ; stipules few, ovate, glandular-serrate. Catkins( the largest 1 inch long) terminal on leafy stalks ; scales spathulate and involvent, mostly pale and either scarcely or distinctly darker-coloured at the apex, clothed with long white hairs ; ovary subulate, glabr ous, on a pedicel as long as or slightly shorter than the long thin linear nectary ; style pale, long and slender ; stigmas long, bifid, spreading. Whilst in both of these forms the herbacea element is very per- 426 DR. F. BUCHANAN WHITE'S ceptible in all the parts, the derivation from Zanata is shown chiefly in the catkins, though it has left its impression in the other organs as well. I have some wiilows (from the same locality) under observation which may prove to be forms of Salix Stephania more nearly related to S. Janata than the two described above. In Juste Botan. Jahresber. for 1885 a bybrid willow found by Strömfelt (in Iceland P) is referred to S. herbacea x lanata, n. hybr. f. a. pubescens, Lundstr., and B. glabrata, Lundstr. 18. Sarix Lapronvum, L. Out of S. Lapponum, as in the case of many other willows, several supposed species were manufactured in the early times of salicology. Most of these have now been abandoned, though in the British lists three are still retained as varieties, viz. :— a. arenaria (L. ex p.), b. Stuartiana (Sm.), and c. pseudo-glauca, Syme (— S. glauca, Sm.). Regarding these, Boswell-Syme says that in his opinion they are scarcely distinguishable; J. D. Hooker characterizes a and b as only slight varieties, and doubts S. glauca, Sm., being the same species; and Babington thinks that S. glauca, Sm., is probably not a native, and that it is hardly the same as a and 5. These varieties are not admitted by the Continental sali- cologists, who, however, distinguish some other forms. Anders- son admits one only—Aelvetica (Vill.), whieh Wimmer considers to be a distinct species more allied to S. glauca, L., than to S. Lapponum. "Wimmer gives two varieties—b. marrubiifolia (Tausch), a very woolly form; and e. Daphneola (Tausch), a narrow-leafed glabrous or almost glabrous plant. Of the British varieties, a. arenaria and b. Stuartiana have no claim to be maintained as distinct. b differs from a only in the greater woolliness of the leaves, and is probably the same as Wimmer's var. b. marrubiifolia, which also cannot well be retained. As regards 8. glauca, Sm., the case is different, as will be explained in treating of the var. or subspecies helvetica (Vill.). But whilst no varietal names can well be given, it is not to be denied that S. Lapponum shows a considerable range of variation. The leaves of our Scottish plant vary both in shape and size and in the amount of pubescence. In shape they range from narrowly- lanceolate to roundly-ovate, some of the latter simulating the form REVISION OF THE BRITISH WILLOWS. 427 argentea or arenaria of Salix repens, whence arose Linné's con- fusion of the name arenaria. Whilst the leaves are usually well- clothed with pubescence, specimens occur which approach a gla- brous condition, but in the absence of fruit cannot be identified with Wimmer's var. Daphneola. The catkins vary a good deal in size, and the capsules both in shape and in the structure of the styles and stigmas. In flowers, which seem to be about the same age, and hence comparable, one set has lanceolate-subulate subacute capsules, distinctly pedicelled, but with the pedicei usually shorter than and very rarely as long as the nectary; whilst another set has ovate-conie smaller sessile capsules with longer styles. In other respects the plants do not present much difference, and intermediate forms connect the two seta. Though S. Lapponwm is most usually a truly alpine species, rarely descending (in Central Scotland) below an altitude of 2000 feet or thereabouts above sea-level, it does occasionally occur in the low ground. On the south side of the Ochil Hills, in Perthshire, a few bushes (discovered by my friend Mr. W. Martin) grow at the ed ge of a field at an elevation of only 700 feet above sea-level. So far as I know, S. Lapponum has not been found in any other part of that range of hills, which, moreover, are as regards that neighbourhood almost devoid of alpine plants. In this locality (which I have visited) it grows with S. pentandra, and could not, to all appearance, have been brought down by water from any sufficiently high altitude. As mentioned in the ‘Student’s Flora’ and elsewhere, S. Lap- ponun has also been found near Edinburgh. The recorded locality 18 Colinton (sometimes erroneously written Collingtou) Woods; and in Edinburgh University Herbarium is a specimen (with 9 catkins) collected there by Greville in 1824. More recently have seen, in the same herbarium, other specimens, collected many years ago, from Craigcrook and Dalkeith Woods, both of which places are near Edinburgh. Whether the species has any claim to be considered native in these three localities, or whether it still occurs there, local botanists must decide. In England this species has been found on Helvellyn only (B. King, 1880). - Lapponum hybridizes with several other species, and some of these hybrids have been found in Britain. Others probably Occur and should be looked for. 428 DR. F. BUCHANAN WHITE’S Var. or Subspecies Salix helvetica, Vill. Whilst Andersson considers S. helvetica, Vill., to be a variety of S. Lapponum, Wimmer treats it as a distinct species, placing it in his ninth tribe, whereas S. Lapponum is in the fifth. At the same time he says that it must be left to the Swiss botanists to decide whether it is a variety or hybrid of S. Lapponum or a dis- tinct species. S. helvetica combines to a certain degree the characters of S. glauca and S. Lapponum. From the latter it differs chiefly by the leaves being always or finally glabrous above, by the catkins being on distinct leafy peduncles, by the styles being bifid or subbifid, and by the paler scales. One or other of all these cha- racters may occasionally be seen to a certain extent in true S. Lapponum, but yet S. helvetica seems to merit varietal or sub- specific rank. From a comparison of specimens and descriptions, I had already come to the conclusion that Smith's S. glauca (which is not the Linnean species) was the same as S. helvetica, when I noticed that Walker-Arnott had apparently expressed what is practically the same opinion. As a synonym under his variety B of S. arenaria (=S. Lapponum) he mentions S. glauca, Sm., and in the notes he says: “ For our var. B we give no stations because we have no reason to believe it indigenous, ..... it is commonly cultivated . . . and is common in Switzerland, where we believe our var. a does not occur.” Though he is wrong in the latter opinion, yet S. helvetica is evidently the Swiss plant he had in view. Whilst, therefore, there is no doubt that Smith's S. glauca is the same as S. helvetica, Vill., it is not quite certain that it is a British plant. Smith gives as localities *in the Highlands of Scotland. Mr. Dickson. On the Clova mountains; Mr. G. and Mr. D. Don. Hooker." Walker-Arnott says that “Mr. Don's specimens now before us from the Clova Mountains are the same as S. arenaria, E. Bot. ;" and Babington states that Smith's specimens came from Mr. Crowe's garden. On the other hand, there is before me a specimen in Edin. Univ. Herbarium labelled by Winch “ Salix glauca, Ben Lawers.” This agrees with the plant cultivated by Mr. Leefe as $. glauca, Sm., “ re- ceived many years ago from the Cambridge Botanic Gardens,” and both are, without doubt, referable to S. helvetica. If Winch’s specimen really came from Ben Lawers (and I see no REVISION OF THE BRITISH WILLOWS. 429 reason why it should be doubted), then Salix helvetica must be admitted as a British plant. At the same time it is desirable that it should be rediscovered. As regards the figures of Smith's glauca, Wimmer condemns, on the authority of Tausch, Eng. Bot. t. 1810, as representing S. glauca, L., and though he cites Sal. Wob. t. 68 under that species, he remarks that it is not good, which, as it illustrates a form intermediate between S. glauca, L., and S. Lapponum, it could not well be expected to be. Smith cites Haller Hist. t. 14 as well representing his glauca, and Wimmer cites the same under S. helvetica, which affords further evidence of the identity of Smith's species. Smith himself, however, gives S. helvetica, Vill., as a synonym of S. arenaria, L. (— S. Lapponum). In Europe, whereas S. Lapponum is both arctic and alpine, S. helvetica is alpine only (Switzerland, Tyrol, and Dauphiné). X SALIX AURITA-LAPPONUM, Wimm. In Edinburgh University Herbarium is an old specimen of a willow collected, I think, by the late Prof. J. H. Balfour, and labelled o Colinton, Edin.,” but without a date. It has one 9 catkin, and is undoubtedly a hybrid between S. Lapponum and S. aurita, both of which species occur, or used to occur, at Colinton, although, as mentioned under that species, it is doubtful whether S. Lapponum is native there. Evenif it were planted, that is no reason either for or against the hybrid being of spontaneous origin. There is also a leaf-specimen, collected near Craigerook In 1832 by the same botanist, which, from the structure of the leaves, I refer to this hybrid. As stated under S. Lapponum, that species has also been found at Craigcrook. Andersson combines the hybrids of S. Lapponum with A Caprea, S. cinerea, and S. aurita, under the name S. Lesfadiana, Hartm., and places S. aurita-Lapponum, Wimm., as b. opaca, 2° subaurita, Like other hybrids, S. awrita-Lapponum exhibits various com- binations of the characters of its parents, but is best known from Its close ally S. cinerea-limosa, Lestad., by the rugosity of the leaves, especially the young ones. It occurs in several parts of Europe (Lapland, Sweden, Silesia, Switzerland), and ought to be found in the Scottish Highlands, where the parent species not unfrequently grow in sufficient proximity. LINN. JOURN.— BOTANY, VOL. XXVII. 26 430 DR. F. BUCHANAN WHITE'S A willow, in Kew Herbarium, collected by Lightfoot, but without locality, date, or name, seems to be Salix aurita-Lap- ponum. X SALIX CINEREA-LIMOSA, Lestad. (S. Lapponum x 8 cinerea.) A specimen, in Edinburgh University Herbarium, labelled “Salix cinerea, Carlowrie, 1838” by, I think, J. H. Balfour, is evidently a hybrid of S. Lapponum, and is very like S. aurita- Lapponum. From the densely black-pubescent twigs, the abundant rusty cinerea-like hairs on the underside of the leaves, and the absence of the rugosity of aurita, it is most probably a hybrid with S. cinerea rather than with S. aurita. The specimen has numerous 9 catkins, but only young leaves. Carlowrie is near Edinburgh, and though there is no record of S. Lapponum having been found, it may have been planted there as well as in the other places in the district. The hybrid itself is not at all likely to have been introduced, but rather to have had a spontaneous origin. S. cinerea-limosa, Liestad. (S. cinerea- Lapponum, Wimm. ; 8. Lestadiana, Hartm., 9. opaca, 1° subcinerea, And.), seems to be of rare occurrence, recorded from North Scandinavia only. It is not so likely to occur on our hills as S. aurita- Lapponum. X SALIX SPURIA (Schleich.), Willd. (S. Lapponum x 8. Arbus- cula.) Since S. Lapponum and 8. Arbuscuca are (with the exception of 5. herbacea) probably the two commonest of our alpine willows, they, as might be expected, eross with each other, though, on the Coutinent at least, apparently not so frequently as might be supposed. ‘Tue hybrid, though at first at least not recognized as such, has been known for a considerable time under the name of S. spuria, Schleich., Willd.; but Wimmer, according to his custom, altered the name to S. Lapponum-Arbuscula. It occurs both in Switzerland and the Tyrol, but, judging by its absence from many herbariums, seems to be rare. In Britain, though it 18 about Seventy years since it was first gathered, it has nol, until quite recently, been identified, having usually been passed over as a form of S. Arbuscula. REVISION OF THE BRITISH WILLOWS. 431 Although I have not been able* to examine European examples of Salix spuria, I have before me a series of willows which are beyond doubt hybrids between S. Lapponum and S. Arbuscula, and which must therefore be called S. spuria. Though, as in the case of other hybrids, some specimens show more affinity with one parent than with the other, the examples I have seen are all tolerably intermediate in character. Judging from the words of Andersson and Wimmer, and from the names given to it by Seringe and Gaudin, the Continental plant seems to have a greater resemblance to S. Lapponum than has the British one. Andersson says that it is so similar to that species that it can scarcely be distinguished, except by the shorter and more compact catkins surrounded by a few leaves, but especially by the much harder and more glabrous leaves finely serrated on the margin. Wimmer describes it as approaching S. Lapponum in the size and clothing of the leaves, and shape and length of the catkins; but to 9. Arbuscula in the glandular-serrate leaves, Which are somewhat shining above, in the narrower ferruginous scales, and in the structure of the stigmas. The Scottish plant, on the other hand, is more likely to be taken—as indeed it has been—for S. Arbuscula than for S. Lapponwm. In size it is nearer the former than the latter, but otherwise it shows a combination of the characters of both. From S. Arbuscula it may be distinguished generally by the duller colour of the leaves, which are more or less, but never excessively, pubescent—the pubescence combining the silkiness of Arbuscula with the woolliness of Lapponum, by the finer and more scanty serration of the leaf-margins, by the longer shape of the capsules, longer styles, and usually narrower scales darker at their tips; and from $. Lapponum by the firmer and more shining leaves, which are more nearly glabrous and have more or less serrate margins, by the smaller catkins with short leafy Peduncles, and by the short stigmas. The characters, however, are more readily seen than described. I have seen British specimens from the Breadalbane Moun- tains, In the former it seems to have been first gathered by Sir . D. Hooker, since in Mr. Hanbury's “ Boswell Herbarium " is * Since this was written I have received (through the kind services of Mr. A. Penn ett) Scandinavian specimens of Lapponum x Arbuscula, which prove to be tentical with some of our Scottish examples. 262 482 DR. F. BUCHANAN WHITE’S a specimen (without date and without flowers) labelled S. Arbus- cula. Of recent years it has been found in several places in Breadalbane, with both d and 9 catkins, by Messrs. Meldrum, Haggart, and Brebner, and by myself. Group 9. Nrrtnpts 14. SALIX Myrsinrres, L. Considering how, on the whole, well-marked this species is, great discrepancies occur in the various descriptions. Andersson says that the ovary is sessile, with the nectary passing its base; Wimmer (with whom Grenier agrees) that it is shortly pedi- cellate with the nectary reaching its body ; W. J. Hooker that it is sessile (in S. procumbens nearly sessile) ; Walker-Arnott (for both “ species") that the pedicel is usually as long as or at length longer than the nectary ; Boswell-Syme that the pedicel is about as long as the nectary ; Babington that the ovary is sub- sessile; and J. D. Hooker that it is distinctly pedicelled. Then as regards the shape of the ovary, Andersson describes it as ovate-conic rostrate (in B. Jacquiniana, thickly ovate and scarcely pedicellate) ; Wimmer, conic-oblong, always obtuse 12 a, more slender and often somewhat acute in B. Jacquiniana; Koch, lanceolate-acuminate from an ovate base, at first sessile and hairy, at length very shortly pedicellate and glabrous; Grenier, ovoid shortly conic; W. J. Hooker and Walker-Arnott, lan- ceolate; Boswell-Syme, lanceolate-conieal or conical-subulate; Babington, ovate-subulate. (By Andersson the d catkins are described as yellowish, the anthers at length becoming blackish; but Wimmer more correctly says that the filaments are purplish, and the anthers purple- violet.) From the descriptions of the ovary, it would appear that British botanists attribute a longer pedicel to the species than the Continental botanists do, and that on the whole they are inclined to describe the shape of the ovary as lanceolate. The result of the examination of a number of specimens shows that there is a considerable range of variation in the lengtb of the pedicel and shape of the ovary, without, however, any notably extreme forms. The Central-European plant seems usually to have a shorter capsule and pedicel than the North-European. Seottish examples are more (though not altogether) in accord- REVISION OF THE BRITISH WILLOWS. 433 ance with the latter than with the former. In them the pedicel is always (so far as I have seen) present, and is often twice as long as the nectary, though it varies in length even in the same catkin. In addition to these variations in the ovary, the structure of the style and stigmas, the shape and size of the catkins and of the leaves, as well as the habit of the plant, are all subject to modification, so that as regards the varieties serrata and arbuti- Jolia of British lists there is no reason for which they deserve to be retained. Nor is it by any means evident that Salix procumbens, Forbes (now considered by most botanists to be a form of S. Myrsinites) has any characters to warrant its retention. Walker-Arnott maintained it as a distinct species chiefly on account of its elongate catkins ; but, as Boswell-Syme points out, this is not à character of any constancy in S. Myrsinites, nor are there any sufficiently distinct characteristics in the habit of the plant, the shape of the leaves, or the structure of the style. x SALIX WanrENBERGIL, And. (S. Myrsinites x S. nigricans.) In the * Prodromus! Andersson uses the name S. myrsinitoides, Fr., for the willow which Wimmer calls S. Myrsinites-nigricans ; but in Blytt's «N orges Flora,’ pt. ii. 1874, he alters the name to 8. Wahlenbergii, And., and gives as synonyms S. punctata, Wahl., 8. myrsinitoides, Fr., S. nigricans *borealis B. punctata, Hartm., and 8 Myrsinites-nigricans, Wimm. o. , Both Andersson (in the * Prodromus’) and Wimmer (in his Salices ’) appear not to have been very well acquainted with the Species, as their descriptions do not altogether agree, and neither was sure of the distribution. From the ‘ Norges Flora,’ however, it would seem that Andersson had become more familiar with the plant, since the description is greatly amplified. After remarking that, through the hybridization of the very variable Myrsinites With the still more variable nigricans, a great number of intermediate forms occur, he proceeds to notice three chief modifications, namely, a. subnigricans, b. coriacea, and e. sub- Myrsinites, Since, however, these forms pass imperceptibly into each other, it seems scarcely worth while retaining these names as those of distinct varieties. In its best form S. Wahlenbergii combines the characteristics of its Parents, deriving from Myrsinites the rigidity, glossiness, 434 DR. F. BUCHANAN WHITE'S and in part the venation of the leaves, the often erect leafy- peduncled catkins, and the structure and colour of the style and stigmas; and from nigricans the somewhat tomentose twigs and leaves, the greater thinness of the latter, and their greater tendency to become black in drying, the often longer petioles, and the often longer pedicels of the capsules. But, as Andersson remarks, some specimens are very difficult to separate from nigricans, and others pass imperceptibly into Myrsinttes, and thus it is often by no means easy to determine some of the less intermediate examples. In Britain Salix Wahlenbergii is possibly not very much rarer than S. Myrsinites itself, since it occurs in many of the localities of the latter, though it has been passed over as Myrsinites and sometimes as nigricans. I have seen good intermediate speci- mens from Perthshire, Forfarshire, and Aberdeenshire—first gathered in the latter two counties by J. H. Balfour, Greville, and their contemporaries ; and in addition many examples gradually passing into either Myrsinites or nigricans, and often very difficult to separate. Neither Wimmer nor Andersson seem to have been acquainted with the d plant, though the former quotes of the catkins “ d elongati, graciles." The g has, however, been found on Ben Laoigh by the Messrs. Groves, and on Ben Heasgarnich by myself. These g plants have leaves near Myrsinites, catkins short, thick, on leafy peduncles, and anthers subglobose and yellow, this yellow colour being the chief point in which they differ from Myrsinites. ; Salix MacNabiana, M acGillivray (‘ Edinb. New Philos. Journ. ix. p. 335, 1830), is possibly this hybrid, but MacGillivray's own herbarium specimens (now in the possession of Dr. Roy, of Aberdeen)—from the Corrie of Loch Kandor, where the hybrid is common—are very close to, if not identical with, Myrsinites, though they are too poor for absolute certainty. A hybrid between $. Myrsinites and S. phylicifolia probably occurs in Britain, but I have not yet seen exampies sufficiently certain to enable me to include it in the list. X SALIX SAXETANA, n. hybr. ($. Myrsinites x S. aurita.) A willow gathered by myself some years ago, and more recently by Messrs. Groves, on Ben Laoigh, Perthshire, puzzled REVISION OF THE BRITISH WILLOWS. 435 me for a long time. I have found quite lately, however, several bushes which, by their range of variation, indicate pretty clearly the affinities. The plant seems, with little doubt, to be a hybrid of Salix Myrsinites and S. aurita, both of which occur on Ben Laoigh, but is so altered that at first sight it does not sugges either of the parents. As yet I have seen 9 catkins only—from at least three bushes, but I think that the d also occurs *. The following description is made from a form tolerably inter- mediate in its characters. It makes a low thick-stemmed bush, with olive-green (turning brown when dried) slightly shining twigs, which are at first somewhat pubescent, but soon become glabrous. Leaves rather large (2 x 1 inches), especially at the top of the shoots, obovate, with a more or less oblique short point ; margin wavy, coarsely erenate-serrate ; upper surface dark shining green, with (when living) the veins slightly im- pressed ; under surface dull pale green or subglaucous, with the primary veins raised; young leaves slightly pubescent, soon becoming glabrous, except on (to a greater or less extent) the midrib and rather long petiole. Stipules very small, 1-cordate, glandular. Catkins moderate in size, erect or spreading, on long or short peduncles which are furnished with several small leaves Whose axils have buds in them; scales narrow, pointed, with the upper half or third fuscous black ; capsule rather small, conical- subulate, clothed with dense white hairs, and on a pedicel which is three to four times as long as the small quadrate yellow néctary; style rather short, bifid to, or nearly to, the base ; stigmas rather short, bifid, spreading. Specimens from another bush show much the same characters, * Since this was written, the supposed ¢ plant has flowered in cultivation. The following description was taken from a living specimen :—Catkins ovate- oblong, lateral, on a peduncle with about three leaves which are as long as the catkin. Peduncular leaves minutely stipulate, glabrous, pale dull green, with the chief veins yellowish green and impressed above ; below clothed with rather straight hairs on the midrib, chief veins, and apical margin, dull paler green, with the principal veins yellowish and raised ; veins translucent, those Which arise from the midrib becoming subparallel to it; margin finely glandular- serrate. Scales broad, greenish white below, the upper third or fourth becoming black (the scales at the top of the catkin are red between the pale and black Portions) ; upper part sparingly hairy with long whitish hairs; apex usually emarginate or erose, Filaments pale, glabrous. Anthers (with cells unequal) subquadrate, at first tinged with red, becoming yellow. Nectary rather small, thick, oblong, usually entire, but sometimes divided into two or more irregular Pleces, green or yellowish-green, The date of flowering was the end of May. 436 DR. F. BUCHANAN WHITE’S but the capsule is less subulate and more cylindrical and obtuse, and seems to show below the hairs a trace of the scaly pubescence which is sometimes characteristic of Salix Myrsinites. A third bush, while practically the same as the others, shows rather a greater affinity with S. aurita. The twigs are more slender ; the young leaves more pubescent; the margins of the leaves more serrate; the catkins smaller, with narrower, more ferruginous scales; the capsule more like that of aurita, with usually an almost obsolete undivided style, and very short, bifid, and erect stigmas. It will be seen from the foregoing that, though the leaves have to a great extent retained the outline of those of S. aurita, they have lost the rugosity characteristic of that species, though in their young state they show in part the pubescence. From S. Myrsinites the shining upper surface has been derived, but the substance is thinner than in that plant. The capsule struc- ture shows varying affinities with both species. Besides the Ben Laoigh examples, I have seen specimens collected in Clova, Forfarshire, by the Rev. W. R. Linton, which—although there are no catkins—I believe to be another form of S. saxetana. In shape the rather small leaves recall both S. Myrsinites and S. aurita. They are roundish- or oblong- obovate, with short more or less oblique points; green but (at least when dried) only slightly shining on eithersurface; margins fnely erenate-serrate; both surfaces sparingly pubescent—the pubescence becoming scantier, but scarcely altogether vanishing in the older leaves ; young leaves rugose, with impressed veins above and raised veins below; older leaves flatter above, but with conspicuous pale (when dried) raised veins below; petioles rather long ; stipules present, but small. I think that, so far as the leaves go, there can be no doubt as to the parentage of this plant. X SALIX SERTA, n. hybr. (S. M. yrsinites x S. Arbuscula.) A specimen in Mr. F. J. Hanbury's ** Boswell Herbarium," labelled “ Salix Arbuscula. Breadalbane Mts., Lyon,” seems almost certainly a hybrid between Myrsinites and Arbuscula. The leaves recall both species, and are rather small, slightly obovate, shining above, and dull and glaucous below, finely crenate-serrate on the margins, quite glabrous (except when very young, when they have the pubescence of Arbuscula), and veined REVISION OF THE BRITISH WILLOWS. 497 in the style of Myrsinites rather than Arbuscula. The catkins (9), which are borne on terminal and lateral, rather long, leafy peduncles, are 14-2 inches long, stout and cylindrical; scales small and rather narrow, very hairy ; capsules sessile or nearly so, oblong-conical, obtuse, coloured like those of Arbuscula, and hairy ; style slender, rather short, with the apex bifid; stigmas short, slender, bifid. The plant is quite intermediate between its supposed parents. Here also may belong a scrap in the same herbarium labelled " Salix prunifolia. Breadalbane Mts. J.D. Hooker.” If it does, itis nearer Arbuseula. The small leaves are rather intermediate ; the capsule more slender than in the above; the style very short, and the stigmas very short and thicker. Under S. Arbuscula, Walker-Arnott describes a var. B from Ben Lawers, with *leaves (broadly or roundish ovate, promi- nently veined above) green, but scarcely shining on both sides." Of it, he says that it is precisely intermediate between Arbuscula and Myrsinites, and may perhaps be a hybrid. It was found only once. 15. SALIX HERBACEA, L. S. herbacea varies in many small particulars, such as size, shape and extent of serration of the leaves, downiness of the young shoots, Ze, The pedicel of the capsule varies a little in length, and in being glabrous or silky, the silkiness sometimes extending to the base of the capsule. The style also varies in length. Occasionally the capsule has a few hairs on it, and more rarely well-developed lines of hairs. The most extreme state of this form which I have seen is in a plant, collected by Dr. Greville m Corrie Kandor in 1830, which has, moreover, the style bifid to the base. A less extreme form I have found near Glen Tilt. Andersson describes a var. subpolaris whose capsule has lines of hairs, but it differs from the type in also having entire leaves. "BAL Grauami (Borr.), Baker. (S. herbacea X S. phylici- Jolia?) Under 8, Myrsinites Sir J. D. Hooker has the following note:—“ 8 Grahami, Borr. MS., is only known from Q speci- mens, cultivated in the Edin. Bot. Garden, said to have been brought by Prof. Graham from Frouvyn in Sutherlandshire. 438 DR. T. BUCHANAN WHITE'S It appears to me to be a form of Saliz Myrsinites, with smaller catkins, paler scales, and a perfectly glabrous capsule with a rather long very silky pedicel; and not allied to S. polaris or herbacea. Syme suggests it to be a hybrid between herbacea and nigricans or phylicifolia; and Nyman a subsp. of S. retusa, L. The Engl. Bot. figures of the ovary and scale are very incorrect. A similar plant occurs in Muckish Mt., Donegal.” Having examined living specimens of both the Scottish and Irish plants, I think that Boswell-Syme is probably correct in his suggestion of the parentage; but I cannot agree with Leefe in considering that the Irish plant “is as nearly as may be identical” with the Sutherland form. In fact, I suspect (although, till it is proved by experiment, it is only a conjecture) that whilst S. herbacea is one of the parents of both, S. phylici- Jolia is the other parent of the Scottish plant, and S. nigricans of the Irish. If this be the case, the former must bear the name of S. Grahami, and the latter that of S. Moorei. The published descriptions of neither of them are absolutely correct, but it will be sufficient to indicate the points in which the two differ. X S. Grahami (Borr.), Baker. Young leaves brighter and more shinin g, rather broader, less pubescent, the pubescence being more of the nature of that of S. phylicifolia, but not markedly so. Scales involvent, broader upwards, obtuse, subemarginate at apex, hairy at the base and ciliate on the margin. Pedicel of the ovary silky-hairy, the pubescence spreading more or less over the base of the ovary, Which is otherwise glabrous. Style rather stouter ; nectary oval- oblong. X S. Moorei (* Watson, L. C.”). Young leaves duller, narrower, and more hairy, the pubescence resembling that of S. nigricans. Scales much longer in pro- portion, narrow oblong, subobtuse at apex, and more hairy. Pedicel of ovary apparently longer, glabrous or slightly hairy; ovary more or less pubescent towards the apex. Nectary linear- oblong, about as long as the pedicel, finally much shorter. The most apparent differences between the two lie in the very differently shaped scales and in the pubescence of the ovary. Boswell-Syme describes the peduncle of the catkin of the Suther- REVISION OF THE BRITISH WILLOWS. 439 land plant as glabrous, but, in all the specimens I have seen, it, as well as the rhachis, is distinctly pubescent. Stipules were unknown him and are certainly rare, but when present are small, more or less narrowly ovate, and toothed. Possibly they are more fre- quent in Salix Moore? than in S. Grahami. In some wild specimens of S. Moore? (kindly lent me by Mr. F. Moore) there are ripe capsules. These are subulate from an ovate base, and about } inch long. The facies of the wild specimens favours the theory that nigricans is one of the parents. The late Dr. Moore, in recording the occurrence of the Irish plant, says that it is the same as specimens of S. Grahami from the Sow of Athole in Perthshire. I have not been able to learn anything about this Sow of Athole plant; but in Borrer’s Her- barium at Kew is a plant, placed under S. Arbuscula, which was collected on that hill by Mr. J. Ball. Its condition is bad, but it seems not to be S. Arbuscula, and may be a form of the Sutherland Grahami, though not agreeing with it in the scales or leaves. X SALIX sIMULATRIX, n. hybr. (S. herbacea x S. Arbuscula.) Under this name I place specimens of four plants from the Breadalbane Mountains, which, though unlike each other, seem to be probably hybrids of S. herbacea with 5S. Arbuscula. They all however, require further investigation. These plants and their characters are as follows :— l. From Coire Dhubh Ghalair (J.Brebner). Nearer S. Arbus- cula than S. herbacea. From the latter it derives its habit, slender arcuate branches, roundish oval thinner leaves, and pseudo-terminal catkins (i. e. at the end of a branch and sub- tended by a leaf, but with a bud between the leaf and the peduncle); from S. Arbuscula it has the thicker trunk and more oval leaves, dull glaucous green below and with the smaller Veins less prominent. Whilst bearing a strong resemblance to Arbuscula, it is not exactly like any of the numerous specimens which I have seen. Compared with examples of the latter of the same age, it may be distinguished by the leaves being thinner, more shining above, and roundish oval in shape; by the habit and slender branches ; and by the position of the catkins. There 18 only one eatkin on the specimen, and that not in good con- dition. The capsules, whilst resembling those of Arbuscula, 440 DR. F. BUCHANAN WHITE’S seem to be a little longer in proportion to their size than in that species. 2. A specimen in F. J. Hanbury’s “ Boswell Herbarium,” on a sheet with the label “ Salix prunifolia, Breadalbane Mts., J. D. Hooker.” This has no flowers, but is in habit nearer herbacea than No. 1, with which in the leaves it quite agrees. 3. Meall Dhuin Croisg (W. Barclay and R. H. Meldrum). In habit and leaves near herbacea, but in its catkins nearer, appa- rently, Arbuscula. I once thought that this might be a form of S. Moorei nearer S. herbacea, but now I suspect that S. Arbuscula is more likely to be one of its parents. The catkins are lateral on leafy peduncles, and are moderately long. The capsules are pubescent, intermediate in shape between those of Arbuscula and herbacea, and pedicellate, with the pedicel about as long as the long linear nectary. The style is of medium length, and the stigmas rather short, thick, and cleft. The leaves much resemble those of herbacea, but are not quite identical. It is beyond doubt a hybrid of herbacea, but whether with Arbuscula is a little uncertain. 4. A specimen in the Edinburgh University Herbarium, labelled “ Salix vaccinifolia, Craig Chailleach, Perthshire; Dr. Hooker.” This has young catkins and young leaves only. The leaves seem essentially the same as those of Nos. 1 and 2. The catkins are lateral and terminal, on leafy peduncles, very small and subglobose. The scales are glabrous on the back and ciliate on the margins, and very similar to those of Aerbacea. The ovaries are like those of Arbuscula, but the style and stigmas like herbacea. Larger and older specimens are desirable. X SALIX SOBRINA, n. hybr. (S. herbacea x S. Lapponum.) A morsel of a willow gathered on the east side of Ben Chat, in Athole Forest, Perthshire, by Dr. Roy, of Aberdeen, and by myself, had long been a puzzle to me (especially as it had no catkins) till I came to examine a plant found by the Rev. E. S. Marshall in Glen Fiagh, Clova, Forfarshire, in 1888. These latter specimens, being in better condition and having 9 catkins, show that it is a hybrid between S. herbacea and S. Lapponum. The examples which I have seen are so intermediate between the parent species, that they are not strikingly like either of them. In size the plant is like herbacea, but in habit more like Lapponum, especially in the comparatively (to the size) stout, REVISION OF THE BRITISH WILLOWS. 441 knotty, tortuous, reddish-brown glabrous branches. The leaves are small, oval, at first woolly on each side, but becoming sub- glabrous, and slightly shining on each surface ; margin obscurely serrate; base rounded; tip (sometimes twisted) subacute and subcartilaginous, as in Lapponum; veins and margin pellucid ; veins finally raised and reticulate on the upper surface; lower surface lineately veined and reticulate with raised veins. Catkins short, few-flowered, lateral on long leafy peduncles ; scales large and broad, obovate, subtruncate, brown with darker tips and long white hairs; capsules sessile, white-woolly, at first conical, then subulate-conical; style moderate, about as long as the rather thick entire or bifid suberect stigmas; nectary long linear. Young shoots pubescent. A willow from Clova (Dr. Greville, 1824), in Edinburgh University Herbarium, may belong here, but if it does it is nearer herbacea than the above. It has leaves only. Since this description was written I have received specimens of A Lapponum-herbacea (no authority for the name, and S. alpestris, And., given as a synonym) from Sweden. These are, to all intents, the same as the Scottish plant, though rather more luxuriant, « Alpestris" is the name adopted by Andersson (in the * Norges Flora’) for the hybrid between 5$. herbacea and S. glauca, and is placed with four other, variously composed, hybrids under 5. norvegica (Fr.), Aud. I have also seen a more numerous series of specimens col- lected in Clova by the Messrs. Linton, as well as what appears to be another form (collected by the same botanists) from Craig- na-dala Beg, Aberdeenshire. These specimens exhibit various degrees of combination between the supposed parents. X BALIX MARGARITA, n. hybr. (S. herbacea X S. aurita.) In Messrs. Groves’ herbarium is a very curious willow, found by them on Craig Loigste, on the south side of Ben Challum, in Perthshire, in 1855; and growing in the Edinburgh Botanic Garden are two willows (for specimens of which I am indebted to Mr. Lindsay, the Superintendent) which, though not identical with the Messrs. Groves’ plant, are apparently another form of the same species. The Edinburgh Garden plants were found in 1875, by the late Prof. Dickson and the late Mr. J. Sadler, near ly ndrum, and probably both on Ben Challum, since a specimen 442 DR. F. BUCHANAN WHITE's sent by Mr. Sadler to Mr. Leefe, and now in Kew Herbarium, has Ben Challum given as the locality. Mr. Sadler describes the wild plant as having “a dwarf procumbent habit;" and Messrs. Groves’ specimens show this, but the cultivated specimens sent to me look as if they had become more upright. The following is a description taken from living specimens of ihe plant found by Mr. Sadler, and now in cultivation :— Twigs divaricate, slender, straight or subflexuose, purple-brown, glabrous, shining; older bark brown or very dark olive-green, roughened but shining; youagest shoots with white pubescence, when older green, or where exposed to the light dark red-brown or purplish, shining, but with a few hairs. Buds at first pubescent, then glabrous, yellowish-red, acute. Stipules (rare) half-cordate, hairy, glandular-serrate on the margin. Leaves thin, dark green, slightly shining above, paler and dull below, at first densely woolly with brownish-white hairs, at length almost or quite glabrous; roundish; base rounded, subcordate, or unequal and slightly cuneate; tip with a twisted point, or truncate, or subemarginaie ; upper surface concave ; margin thickened and incurved serrate-crenate, with incurved glandular teeth ; upper surface rugose from the impressed veins, under surface with raised reticulate veins, smaller veins pellucid ; petioles rather long and slender; the largest leaves li inch long and broad, but most of them much smaller; vernation involute. Catkins ( 9) small (less than 4 inch long), dense-flowered, on lateral leafy peduncles of about the same length as the catkin; peduncle-leaves 2-4, ciliate, with buds in their axils, and stipulate ; peduncle and rachis downy; scales oblong, narrow and long, concave, glabrous on the back, ciliate at the apex, and with a few hairs on the inner surface, greenish yellow, those at the top of the catkin tinged with pink at their tips; ovary conical from an ovate base, sub- obtuse, with coarse woolly white pubescence ; pedicel downy, nearly three times as long as the nectary, which is thin in texture, oblong or widened upwards, entire or cleft once or twice at the apex or nearly to the base; style very short, thick, greenish yellow; stigmas bifid, suberect, as long as the style. In one specimen the uppermost scales are widened upwards, broader and involvent, and have a few hairs on the back. In Prof. Dickson’s plant the characters are much the same. The older bark is duller, the buds are less acute, and the catkin- REVISION OF THE BRITISH WILLOWS. 443 scales more uniformly narrow, the catkins rather larger, and the stigmas spreading. Messrs. Groves’ specimens have often obovate and sometimes oblong leaves, the largest about 1 inch long by ? inch broad, but most of them less than half that size; slender, more flexuous twigs; young leaves less hairy; petioles rather shorter; the solitary catkin more slender, with scales shorter, broadly obovate, and involvent, the pedicel of the ovary rather shorter, and the stigmas larger and spreading. There can, I think, be no doubt that this species is a hybrid between Salix herbacea and S. aurita. The Edinburgh Botanic Garden plants are, in the shape of the scales and in the capsules, nearer aurita than herbacea, to which, in these parts, Messrs. Groves’ specimens are more related. Both sets are, however, intermediate in character between their parents. 16. SALIX RETICULATA, L. In several small points S. reticulata departs from the usual descriptions of this well-marked species, but these variations require no special notice. The margin of the leaf is described as entire, but on the appareut margin (though not usually on the real edge) there is frequently a row of glands which give the appearance of a minute serration. Andersson mentions two varieties—a. typica and p. nivalis (S. nivalis, Hooker), the latter being a remarkable small form which occurs in Iceland, Spitzbergen, and N. America. The Var. a. typica is divided into :—1. glabra (leaves quite glabrous on each side), and 2. sericea (leaves more or less villous, and the margin towards the base here and there glandularly subserrate). The latter is the more frequent form in limestone districts of S. Europe. Whilst the Scottish plant must be referred to l. labra, I think that the Glen Callater examples retain the hairs on the underside of the leaf (both surfaces are more or less hairy when young) for a longer period than the Perthshire ones do. I have seen two Scottish plants which have been referred to S. reticulata, but which seem to be evidently hybrids of it with other species, They are well worth further investigation. _ One of these plants is in the British Museum Herbarium, and 18 labelled « Salicis reticulate varietas ?, Ben Lawers, Perthshire, 1793, R. Brown.” Unfortunately it has no flowers, but it may be thus described :— 444 DR. F. BUCHANAN WHITE'S In habit like S. reticulata, but stems more slender, rooting, chestnut-brown. Petioles very short, being little, if at all, longer than the bud in the axil. Leaves narrowly oval or elliptic, some of them rather broader beyond the middle, attenuate at the base, flat above and below and glaucous below; midrib and veins slender, primary veins leaving the midrib at a very acute angle and running parallel to it, secondary veins anastomosing ; margin cartilaginously thickened, entire; texture apparently thinner than in reticulata; young leaves with a few hairs below, soon becoming quite glabrous. Largest leaves nearly 2 inches long by 2 inch broad; average size 11x 1 inch. The very short petioles &c. indicate the distinctness of this plant (which may be provisionally called S. sejuncía) from S. reticulata. The other plant referred to is in the Kew Herbarium, and is labelled “ Salix reticulata, L., And." (meaning that Andersson certified it), * Scotland, Herb. Lambert." It is quite a different looking plant from S. reticulata. The stems are straighter than in that species, and more or less woolly. The leaves are less strongly reticulate, quadrately oval, cordate at the base, distinctly and rather closely crenulate-serrate, woolly on each side but more especially above; petiole very short, and very woolly, as is the young shoot. The catkins are lateral ; the scales (not in good condition) seem to be larger ; cap- sule (burst) sessile, glabrous, shaped as in reticulata; style rather long; stigmas rather long, entire or bifid. This apparently very distinct plant I provisionally name S. soluta; but I am unwilling to place either it or the other in the list till they have been rediscovered. X SALIX SEMIRETICULATA, n. hybr. (S. reticulata x 8. ni- gricans ? ) A willow which was discovered at an altitude of 2300 feet on Meall Ghaordie, in Perthshire, by Mr. James Brebner, has a superficial resemblance to S. Grahami, chiefly from the habit of the plant and the shape of the leaves. It seems, however, to be a hybrid between S. reticulata and probably 5. nigri- cans, although it has not much resemblance to either of these species. It may be thus described :— Branches slender, long and trailing; bark fuscous brown glabrous and shining; shoots of the year dull paler brown white-pubescent. Buds red-brown, at first slightly pubescent REVISION OF THE BRITISH WILLOWS. 445 with white hairs, finally glabrous, quadrately-ovate and obtuse. Leaves at first half-folded, the older ones flatter, oblong-orbicular, truncate or subcordate at the base, obtuse or slightly twisted at the apex; margin slightly reflexed, slightly and rather remotely glandular serrate-crenate or nearly or quite entire, youngest leaves less obscurely crenate; surfaces at first, and especially below, more or less furnished with white woolly adpressed hairs, afterwards glabrous above and nearly or quite glabrous below; upper surface dark green, shining, reticulately rugose from all the veins, even in the smallest, being impressed (when dried, how- ever, some of the veins are raised); under surface dull paler green, white-dotted as is also the upper surface, the veins in- cluding the smaller ones somewhat thickened and reticulately prominent, more especially when young; veins numerous, pel- lucid; primary veins forming an acute angle with the midrib, those towards the base of the leaf more approximate; larger leaves about 1X4 inch, smaller 4X4 inch; petiole about À of the lamina in length, grey-woolly above; stipules (rare) minute, roundish, glandular, soon disappearing ; leaves involute in verna- tion. Catkins lateral on peduncles about as long as the catkin, and with three or four leaves with buds in the axils, the leaves similar to the other leaves but smaller, and the upper surface mostly glabrous, narrowly red at and near the apex; catkins ovate, short, compact; rachis stout, woolly ; scales membranous, brownish, clothed with long straight white hairs on both surfaces, broadly spathulate, obtusely rounded and erose at the apex; capsules (large for the size of the catkin) ovate subulate, slightly compressed obtuse (acute in the young ovary), green tinged more or less with red-brown, slightly hairy especially towards the top, but sometimes becoming almost or quite glabrous ; Pedicel silky or glabrous, stout, shorter than the long, linear, Very thin nectary ; style moderate, bifid at apex; stigmas about as long as the style, bifid, erect, but twisted and recurved at the apex *, * The above description was taken from the rather old catkins of the wild Specimens. The following is from a younger catkin from a cultivated plant:— les veined, broadly obovate, involvent ; apex truncate or slightly rounded, and sometimes slightly emarginate ; gieenish white at the base, brownish, reddish brown, or red in the middle (the upper scales being the brightest coloured), and black at the apex; upper part of both surfaces clothed with long white hairs, Ovary subulate-conic, shortly pedicellate, both ovary and pedicel with long hairs or somewhat glabrous below ; style stout, yellow, its L INN. JOURN.— BOTANY, VOL. XXVII. 2H 446 DR. F. BUCHANAN WHITE'S In some specimens the midrib and primary veins, as well as the petiole and margin, are somewhat reddish in colour. A willow which Mr. Brebner found on a neighbouring rock to the above much resembles it, but has flatter, more oblong leaves, more closely but shallowly crenate-serrate or almost serrate, more distinctly white-dotted above, less rugose above and less reticulate below, veins less pellucid, and margins less reflexed. I have not yet seen catkins *. In Smith's herbarium is a willow labelled in ink * Salix ellip- tica, nov. sp., Clova Mountains, Mr. Thos. Drummond, Mr. W. Hobertson, 1825," and in pencil *I have this as a rounder- leaved var. of S. Myrsinites.” It has no catkins, but—from memory—is much like S. semireticulata. I think that it has no connection with S. Myrsinites. X SALIX SIBYLLINA, n. hybr. (S. reticulata x S. Lapponum.) In Edinburgh University Herbarium are four small specimens of a willow found by Greville near Loch Brandy, Clova, m 1824. They are labelled “ possibly a pilose state of 5S. reticulata. Though without flowers, these specimens so evidently belong, I think, to a hybrid between S. reticulata and S. Lapponum, that I have ventured to give them a name and a place in this list- The twigs are intermediate (between the parents) in character, rather stout, shining, chestnut becoming grey; young shoots slender, somewhat pubescent. Leaves small (the largest 2X2 inch, but many are much smaller), ovate, ovate-oblong, or gub- rotund, at first woolly on each surface but more especially below, finally nearly glabrous above and slightly shining ; upper surface at first slightly impressed by reticulate veins ; under surface with primary veins raised, the others scarcely or slightly r aised, reticulate; margin incurved, with obscure glands as in reticulata ; apex obtuse, or with a short cartilaginous point as in Lapponum ; base subcordate, rounded, or shortly cuneate; petioles very short. Whilst bearing no striking resemblance to reticulata, it yet base hidden by the scale; stigmas equal in length to the style, thick, bifid, curved, and connivent, yellow. Nectary much longer than the pedicel of the ovary, linear oblong, rather thick, yellowish green. . * A cultivated plant has since this was written produced Q catkins which though not perfectly identical with those of the other plant, are not essentially different in structure. In the structure of the stigmas and scales there is much resemblance to S, Grahami. f : REVISION OF THE BRITISH WILLOWS. 447. shows in the venation, structure of the margin, and shape of the leaves, its relationship to that species. The specimens are too old to show the colour well, but that of the principal veins isin the direction of S. reticulata, namely reddish. From the look of the examples and the dead wood about them, they seem not to have been in robust condition, and probably the plant grows much larger. C. SYNANDRZ. Group 10. PURPURE X. 17. SALIX PURPUREA, L. Whilst both British and Continental salicologists distinguish several modifications or so-called varieties of S. purpurea, they are not altogether agreed as to what these are or as to the names they should bear. In Britain the “ species” into which Smith and Borrer divided 8. purpurea are still retained as varieties of that plant. These are Lambertiana, Sm., Woolgariana, Borr., and ramulosa, Borr. (sometimes included in Woolgariana); and according to Babing- ton, S. Helix, L. Andersson has, in addition to the type (a gracilis, G. and Gr. =8. purpurea, Sm.), Lambertiana, Sm. (including S. Woolgariana, Borr.), and S. Helix, “ L.” (which is not the S. Helix of British botanists, but seems to be the same as ramulosa, Bort.) Wimmer distinguishes the forms eriantha, gracilis, Lam- bertiana, styligera, sericea, and furcata. As characterizing their varieties, British botanists lay stress Upon the colour of the twigs and upon the form of the stigmas, Points which are almost or quite ignored by the Continental salicologists. That the colour of the bark is not 8 constant character may be learnt from an examination of almost auy living bush, when it will be seen that very frequently the colora- tion depends to a great degree upon exposure to the lisht. oreover, from authentically named specimens it would seem that even British botanists do not attach so much importance to this feature as the descriptions would imply. Thus in Smith's herbarium Lambertiana (from Lambert himself), whose twigs modern writers describe as purplish, has in two of the flowering- specimens dark-coloured, and in the third, pale, bark. 448 DR. F. BUCHANAN WHITES As for the shape of the stigmas, I think that not much reliance can be placed on that supposed characteristic. There remain therefore as primary characters, the form of the leaves, of the catkins, and of the capsules; and following Anders- son's views three chief forms may be distinguished by those who wish to retain varietal names. These are :— l. gracilis (or genuina), with slender catkins, small subovate capsules, and usually narrow leaves. 2. Lambertiana, with larger catkins, ovate-conie capsules, and larger leaves broad throughout (Lambéertiana), or con- spicuously broader above the middle (Woolgariana). 3. ramulosa (S. Helix, And., but since S. Helix, L., is dubious, the use of that name is not expedient), with rather stout capsules, and narrow, more elongate, and more acuminate leaves. After examining an extensive series, the conclusion I come to is that while extreme forms admit of having one or other of these “ varietal ” names applied to them, the various modifications pass insensibly (as Boswell-Syme remarks) into each other, and that there are many specimens of which it is impossible to say with cer- tainty to which variety they belong. Idoubt therefore the expe- diency of retaining any varietal names. Regarding Wimmer's forms (other than those mentioned above), there are amongst British specimens plants that might be referred to eriantha (catkins more hairy) though not in an extreme state, to styligera (ovary with a short style), and to sericea (young leaves more or less woolly). X SALIX RUBRA, Huds. (S. purpurea x S. viminalis.) Since there has never been perfect unanimity amongst salico- logists regarding Linné’s Salix Helix, I have used for the hybrid between S. purpurea and S. viminalis Hudson's common] y accepted name of S. rubra. At the same time it is by no means clear that S. Helix should not be the name to be adopted. Fries and Koch, and at a later period Andersson, thought that S. Helis was only a form of S. purpurea. Wimmer, on the other hand, believing that Linné could not have described under two names such a well-marked species (and one which he bad seen in a living state) as S. purpurea, makes S. Helix a synonym of his S. viminalis-purpurea. Smith and his followers considered S. Helis to be a distinct species; and the later British botanists (with the exception of Babington, who places it under purpurea) agree in thinking it to be a variety of rubra. REVISION OF THE BRITISH WILLOWS. 449 Now, whatever doubt attends the Linnean plant—and it is cer- tainly great,—there is none about Smith's. Both his descrip- tions and his specimens clearly indicate a hybrid between pur- purea and viminalis, though much more closely related to the former than to the latter. Itis to be noted, however, that many herbarium specimens named Salix Helix belong to S. purpurea. As a rule, S. rubra seems to show a smaller range of variation than most other hybrids, and hence its division by Grenier into two chief varieties—viminaloides and purpureoides—has some claim for consideration. But this supposed constancy of form is perhaps more apparent than real; and is due, in the first place, to the fact that in many cases the plants have been derived from cuttings, and not from seeds ; and, in the second, to the leaves of both parent species being of the linear-lanceolate type, and hence not affording so much scope for difference in outline in the hybrid. Var. viminaloides (or rubra genuina)is distinguished by its longer and narrower lanceolate leaves, more or less pubescent below, at least when young, and with the margins more or less revolute ; by the paler and more hairy 9 catkins; and by the longer style and stiginas. The petioles of the leaves are described as longer, but in this and other points there is variability. In some British books the filaments are said to be free except at the base; but in the same catkin filaments almost free or united to any point between the base and the apex may be found. Var. purpureoides (which includes Forbiana, Sm.) has leaves broader at the middle, with flat and more or less serrated margins and more entirely glabrous surfaces ; darker-coloured and less hairy catkins, and more obtuse capsules with shorter styles and thicker stigmas. The leaves are described as more shortly petioled than in viminaloides, and Continental examples often show this, but in authentic examples of the British Forbiana the reverse 18 the case, Andersson adopts Grenier’s division into two varieties ; but Wimmer, who distinguishes several forms, does not: and since ìn places where rubra occurs as a spontaneous hybrid, plants combining the characters of viminaloides and purpureoides may be found, I think that it is not desirable to maintain varieties in this hybrid more than in others. Wimmer’s forms are b. For- byana ; c. sericea, Koch (=S. eleagnifolia, Tausch), with the leaves clothed below with silvery-white hairs—a form which I ave seen in Perthshire; d. macrostigma ; and e. angustissima. These and other described modifications show that though, for 450 DR. F. BUCHANAN WHITE'S the reasons given above, there appears to be a constancy of form in Salix rubra, a series of conditions can be found connecting viminalis and purpurea, the chief grades being sericea, rubra, Forbiana, and Helix (of Smith). A Perthshire plant, otherwise intermediate in its characters, has rather remarkable stamens. Whilst the filaments are usually connate half-way or to near the top, in some of the flowers one or both are divided at the apex—the division being of unequal length, and presenting the appearance of a stamen with 2, 3, or 4 irregular branches. Itis to be noted that as these branches form acute and not obtuse angles with the main filament, this does not seem to be a true case of what Hayne has termed cladostemmy. x SALIX SORDIDA, Kern. (S. purpurea X Ñ. cinerea.) As the name S. Pontederana, Schleich., is, according to A. Kerner, dubious, having been applied to hybrids of S. purpurea with several of the Capree, it appears advisable to use Kerner 8 name of S. sordida for the hybrid with S. cinerea. Andersson, indeed, retains the name Pontederana ; but he makes it include the hybrids formed not only with S. cinerea, but with S. Caprea, S. grandifolia, and S. aurita, since he thinks that there is no sure method of separating them. As, however, he has not united these species, it is scarcely justifiable to unite the hybrids if it is at all possible to distinguish them; and he himself has kept them separate as varieties. Of S. sordida (whose synonyms are A. subpurpurea-cineret, Kern., and S. cinerea-purpurea, Wimm.), Andersson, following Wimmer, describes two modifications, viz.: 1. cinerascens (nearer S. cinerea), and 2. glaucescens (approaching S. purpurea, and probably S. Pontederana of Koch). It is to be noted, however, that in other respects Andersson's and Wimmer's descriptions do not tally in every partieular; and that, moreover, Wimmer's published specimens do not always altogether agree with his de- scriptions, as, for example, in the length ascribed to the pedicel of the capsule. In the Woody Island near Perth, where both S. purpurea and S. cinerea abound, S. sordida appears to be not uncommon. But while forms quite intermediate between the parent species occur; the majority of individuals are in character much nearer s. a- nerea than S. purpurea—so near, in fact, that in the absence of REVISION OF THE BRITISH WILLOWS. 451 flowers, some of them would with difficulty be diseriminated from cinerea otherwise than as slight modifications. As belonging to Salix sordida I put all plants which, however like cinerea they may be, have the filaments of the stamens more or less united to each other. Connate filaments may, according to Wimmer, be distinguished from cladostemmie ones by the branches forming an acute and not an obtuse angle; and one of the best characters of hybridization with S. purpurea is the pre- sence of connate stamens. In these Woody Island plants an extreme degree of connation israre. Frequently the filaments are united fora little way above the base only ; but in the same catkin free, slightly united, and more distinetly connate stamens may all be found—a phenomenon which may be seen in S. rubra and other hybrids of S. purpurea. Other points in which these cinerea-like plants are variable are the shape, colour, and pubescence of the leaves, stoutness and pubescence of the twigs, shape of the catkins, and colour of the anthers. In all these particulars they approach, or recede from, S. cinerea by almost imperceptible gradations. Probably, as Wimmer suggests, the cinerascens modifications of S. sordida have a somewhat different origin from those which are nearer 8. Purpurea, having been possibly produced by purpurea 9 x cinerea 3, or, what seems more likely, by the crossing of sordida with cinerea, The 9 of the cinerea-like forms is a much more difficult plant to distinguish than the d , since of course we have not the assist- ance afforded by the connate filaments. I have indeed found some plants whose catkins are so like those of cinerea, that it is only by the resemblance of their leaves and habit to some d Plants of sordida that they can be referred to that species. In addition to the Woody Island plants, I have seen specimens from a 2 bush found at Dalmarnock, on the Tay above Dunkeld, by Mr. C. M'Inotsh, and have found both sexes on the banks of the Tay below Perth. Another willow from the Woody Island I at present include under S. sordida (as var. rubella), though I am inclined to sus- pect that it may be a cross of S. rubra (with which it grows) and S. cinerea (i. e. S. purpurea x S. viminalis x S. cinerea). In its leaves and habit is is near S. purpurea, but has free stamens. e Very beautiful catkins are small or moderate in size; the Pnopened anthers orange-red, but when burst yellow from the 452 DR. F. BUCHANAN WHITE'S pollen, and finally, when empty, fuscous; leaves oblanceolate, dark green and shining above, very glaucous below, soon quite glabrous, but at first more or less pubescent with brownish hairs ; branches slender and straight, glabrous almost from the first. X SALIX DICHROA, Déll. (S. purpurea x S. aurita.) To this hybrid belongs the willow published by Mr. Leefe (Sal. Exs. iii. No. 59) as “Salix Pontederana ? Schl.," and found by him at Rothbury in Northumberland. S. dichroa is very similar to S. sordida ; but the cinerea element in the latter is replaced by aurita. Mr. Leefe, who describes his plant as *a small shrub with declining branches," has published d specimens only; and these agree so sufficiently well with examples of the hybrid issued by Wimmer and by Kerner, that there can be no doubt about their identity. Mr. Leefe found in the same place a willow which he thought “might be the female;" but the catkins are much deformed, being, in fact, her- maphrodite monstrosities. Of this I have seen one catkin only ; and from it not much can be learnt. The leaves, however, re- semble those of the ¢ plant, but are rather nearer in character to S. aurita. XSaLIx Dontana, Sm. (S. purpurea x S. repens.) Salix Doniana was founded by Smith (‘English Flora, iv. p- 213. 47) on a willow “sent from Scotland, as British, by the late Mr. George Don ;" and though it has continued to be re- tained in our floras, it is probable that the majority of the later botanists had come to the conclusion that its claims to be con- sidered British were very slight. For my own part I had, at the time that this paper was read, decided to omit it from the list of British Willows, because, amongst other reasons, it had not come directly from Don into Smith's hands, but from Borrer, who received it from George Anderson, and in the transmission from hand to hand some particulars of its occurrence might have been accidentally lost. But I have now the pleasure of restoring it to a place in the list, having, during the past summer, found un- doubtedly wild specimens on the banks of the river Tummel, near Pitlochry, in Perthshire. The plants I found were growing with S. purpurea, and in the neighbourhood of S. repens. They are distinctly intermediate in character between the parents. The following notes were taken emer ` mme e REVISION OF THE BRITISH WILLOWS. 453 from living specimens :--Bush about 18 inches high. Bark of the twigs shining brownish green, of the shoots redder and slightly downy. Some of the upper leaves subopposite, after the manner of Salix purpurea. Leaves dull, rather dark green with white hairs above; below dull pale green, with more copious long white hairs; margin cartilaginous, slightly incurved, very finely and remotely serrate, with reddish glands ; veins pellucid, the chief veins impressed on the upper surface and raised on the lower. In shape the leaves are similar to those of the S. pur- purea amongst which it was growing (i. e. similar to S. Doniana, Sal. Wob. t. 85). Catkins ( 9 ) rather old, lateral, on a peduncle having three or four small leaves. Scales ovate spathulate, rounded at the apex, upper half black, with long white hairs. Ovary ovate conic, blunt; style distinct, though short ; stigmas very short, thick, and rather broad, semi-bifid; ovary white pubescent, on a pedicel about as long as the linear pale yellow nectary. Though the 9 plant ouly has been found in Britain, both sexes are known in Central Europe, where S. Doniana is widely distributed. Andersson mentions four forms or varieties, VIZ. a. latifolia, (3. lingulata, y. linearis, ò. leiocarpa. Wimmer de- scribes six forms, but does not give them names. The filaments of the g flowers are connate at the base only, or for a third or half their length. DESCRIPTION OF THE PLATES. Puate IX. s regarding the number of E diagrammati . uu ` € statement of the different view between 1762 and 1886. British Willows held at various periods À k The dates given are those of the publication of the Works Wee representing the views of the period (see p- 334). The shad , columns indicate the number of wumbered “ species,” and are divide into three portions—dark, medium, and light. The dark shaded , ecies ; the medium shade portion shows the number of true sp light shade the number number now recognized as hybrids ; and the ` of those once supposed to be species, but now considered to be oie merely forms or to have no claim to be admitted as British. ^ ` last column is given the estimate for 1890 of the true species ( di shaded) and of the hybrids (medium shaded) which have been found m Britain. LINN. JOURN.— BOTANY, VOL. XXVII. 21 454 DR. F. BUCHANAN WHITE'S Piare X. A diagram to iliustrate the various ways in which compound hybrids (i. e. hybrids in whose parentage are included more than two species) may be formed. Puare XI. A diagram showing the relationship of both Tribes and Species as regards the British simple hybrid Willows. The larger circles represent the Tribes, the smaller circles the Species, and the lines connecting the latter indicate that Hybrids between these species have been found in Britain. The circle containing the Tribe Diandrs has been divided into four sections to show the altitudinal range of the species, since that limits to some degree the number of hybrids. Section I. is strietly lowland, the solitary species in it not ascending above 1000 feet above sea-level. Section II. includes species which, while most common in the low grounds, yet ascend into the alpine region of Section III. Section III. contains the species which rarely, if ever, descend below 1000 feet, and most frequently occur only at a much greater altitude. Section IV. contains two species which properly belong to Section IIL, but which occasionally, though very rarely, descend far below the 1000-feet line. The Tribes Pleiandre and Synandre inciude lowland species only. INDEX. The Tribes are printed in eu 311. CAPITALS, the groups in alics, the species and hybrids of Salix in Roman type, and unnamed hybrids are doubly indented. Capree, 377. voy? Dianpre, 377, Salix Fragiles, 362. Nitidule, 432. Nive@, +2}, Pentandre, 359. Phylicifolie, 395. PLEIANDRA, 347. Purpurea, 447, Repentes, 339. Salix acuminata, Sm., 413, 420, adscendens, Sm., 392. alba, L., 370. alba x decipiens, 355. alba x fragilis, 271. alba x pentandra, 362}. alba x triandra, 354. alupecuruides, Zuusch, 333. alpestris, And., 441. ambigua, Ehrh., 392. amygdalina, Z., 347, 548. angustifolia, Wulf., 391. aquatica, Sm., 381, 389. Arbuscula, L., 392, 410. Arbuscula, Sm., 991. — o Arbuscula x phylicitolia, 417. Arbuscula x herbacea, 439. Arbuscula x Lapponum, 430. Arbuscula x Myrsinites, 426. arenaria, Z., 426. atrocinerea, Brot., 980. aurita, L., 382, 384. aurita X Caprea, 387. aurita x cinerea, 383. aurita x herbacea, 411. aurita X Lapponum, 429. REVISION OF THE BRITISH WILLOWS. Salix aurita X Myrsinites, 434. aurita x nigricans, 409. aurita X phylicifolia, 402, 405. aurita X purpurea, 452. aurita X repens, 392. aurita X viminalis, 414. aurita-Lapponum, Wimm., 429. aurita-nigricans, Heidenr., 409. aurita-viminalis, Wimm., 414. Aurora, Lestad., 392. bicolor, Ehrh., 403. bicolor, Forbes, 403. borealis, Fr., 400. cerulea, Sm., 370. Calodendron. Wimm., 414, 420. campestris, Fr., 400. canariensis, C. Sm., 379. Caprea, L., 385. Caprea x aurita, 387. Caprea X cinerea, 386. Caprea X cinerea X phylicifolia, Caprea X coriacea, 410. Caprea x nigricans, 406. Caprea X phylicifolia, 402. Caprea X repens, 394. Caprea X viminalis, 414. Capree x viminalis, 413. Caprea-dasyclados, Wimm., 420. Caprea-nigricans, Wimm., 408. Caprea-repens, Lasch, 394. Caprea-viminalis, Wimm., 414. capreola, J. Kern., 387. carinata, Sm., 410. cinerea, L., 378. cinerea X aurita, 383. cinerea X Caprea, 386. simer icifoli | 402, x Gaprea X phylicifolia, cinerea X Lapponum, 430. einerea x phylicifolia, 402, 403. cmerea X nigricans, 408. cinerea X repens, 393. cinerea X purpurea, 450. cinerea x rubra, 451. wel X viminalis, 414, ils. c: nerea-Lapponum, Wimm., 430. i erea-limosa, Lestad., 430. vine purpurea, Kern., 450. ane repens, Wimm., 393. Algo Viminalis, Wimm., 414, conformis, Schleich. contorta, Crowe, 347, BAS. coriacea, Forbes, 409. Oriacea x Ca cotinifolia, ane 10. Croweana, Sn., 398. cuspidata, Schultz, 360. aphneola, Tausch, 426. | Salix daphnoides, Vill., 310. dasyclados, Wimm., 4 15, 420. decipiens, Hoffm., 348. decipiens x alba, 355. dichroa, Doll, 452. Dicksoniana, Sm., 399, 412. Doniana, Sm., 452. eleagnifolia, Tausch, 449. elliptica, Sm., 446. excelsior, Host, 374. ferruginea, G. And., 414, 419. ferruginea, Forbes, 414. finmarchica, Willd., 392. firma, Forbes, 407. foetida, Schleich., 410. Forbiana, Sm., 449. formosa, Willd., 411. fragilis, L., 363, 368. fragilis, var. britannica, B. White, 368. . fragilis, var. porcellanea, Baenitz, 5) 350. fragilis X alba, 371. fragilis x pentandra, 360. fragilis x triandra, Wimm., 348. fragilis-triandra, Wimm., 348. Friesiana, And., 391. fusca, L., 391. glauca, L., 426, 428. glauca, Sm., 426, 428. Grahami, Baker, 437, 444. grandifolia, Ser., 378. grisophylla, Forbes, 409. Hegetschweileri, Heer, 399. Lambertiana, Sm., 447 lanata, L., 421. lanata x herbacea, 494. lanata X reticulata, 422. Janceolata, Sm 200. l apponum, L., 426. L T pponun x Arbuscula, 490. Lapponum x aurita, 429. 456 Salix DR. F. BUCHANAN WIIITE'S Salix Lapponum x cinerea, 430. Lapponum x herbacea, 440. Lapponum x reticulata, 446. Lapponum-Arbuscula, Wimm., 430. Lapponum-herbacea, 441. latifolia, Forbes, 406. laurina, Sm., 402. laxiflora, Borr., 403. longifolia, Host, 420. ludificans, B. White, 402, 405. lutescens, 4. Kern., 383. Macnabiana, MacGillivray, 434. margarita, B. White, 441. marrubiifolia. Tausch, 426. Micheliana, Forbes, 418. mollissima, Ehrh., 355, 358. monspeliensis, Forbes, 369. montana, Forbes, 364, 376. Moorei, “Watson, L. C.," 438. multiformis, Doll, 355. myricoides, 404. Myrsinites, L., 432. Myrsinites x Arbuscula, 436. Myrsinites X aurita, 434. Myrsinites x nigricans. 433. Myrsinites x phylicifolia, 434. Myrsinites-nigricans, Wimm., 433. myrsinitoides, Fr., 433. nigricans, Sm., 396, 400. nigricans X aurita, 409. nigricans X Caprea, 406. nigricans X cinerea, 408, nigricans x herbacea, 438. nigricans x Myrsinites, 433. nigricans x phylicifolia, 401. nigricans x repens, 394. nigricans x reticulata, 444. nigricans-repens, Heidenr., 394. nigricans-Weigeliana, Wimm., A01. nivalis, Hook., 443. norvegica, And., 441. obscura, Doll, 419. oleifolia, Sm., 381, 384. palustris, Host, 374. pedicellata, Desf., 378. pendula, Ser., 364, 376. pentandra, L., 359, 365. pentandra x alba, 361. pentandra x fragilis, 360. pentandra x phylicifolia, 361. phylicifolia, L., 396. phylicifolia x Arbuscula, 412. phylicifolia x aurita, 402, 405. phylicifolia x Caprea, 409. phylicifolia x cinerea, 402, 403. phylicifolia x cinerea x Caprea, 402, 406. phylicifolia x herbacea, 437. phylicifolia x Myrsinites, 434. phylicifolia x nigricans, 401. phylicifolia x pentandra, 361. phylicifolia x purpurea, 399. phylicifolia x repens, 395. — az phylicifolia-nigricans, Wimm., 397, 398, 400. plicata, Fr., 392. Pontederana, Koch, 450. ` Pontederana, Schleich., 450. procumbens, Forbes, 433. prunifolia, Sm., 410. , pseudo-stipularis, Lond. Cat., 414, 416. puberula, Döll, 408. punctata, Wahlenb., 438. purpurea, L., 447. — purpurea x aurita, 452. purpurea X cinerea, 450. " purpurea x phylicifolia, 399. purpurea x repens, 452. purpurea x viminalis, 448. ramulosa, Borr., 447. ` Reichardti, A. Kern., 386. repens, L., 389. repens x aurita, 392. repens x Caprea, 31M. repens x cinerea, 393. repens X nigricans, 394. ` repens x phylicifolia, 399. repens X purpurea, 452. repens X viminalis, 391. aq repens-myrtilloides, Wimm., 392. reticulata, L., 443. reticulata x lanata, 429. reticulata X Lapponum, 446. reticulata X nigricans, 444. retusa, L., 438. rhætica, Kern., 399. rosmarinifolia, L., 390. rosmarinifolia, Sm., 391. rubra, Huds., 448. rubra x cinerea, 451. rugosa, Leefe, 414, 41%, 419. Russelliana, Sm., 363, 369. Sadleri, Boswell-Syme, 422. saxetana, B. White, 434. ` Schraderiana. Willd., 390. sejuncta, B. White, 444. semireticulata, B. White, HA sericans, Tausch, 414, 41%. serta, B. White, 436. . sibyllina, B. White, 446. silesiaca, Willd., 378, 38t i. simulatrix, B. White, 439. Smithiana, Willd., 413. sobrina, B. White, 440. soluta, B. White, 144. sordida, Kern., 450. . ei. sordida, var. rubella, B. W Ate, 451 spathulata, Willd., 392. speciosa, Host, 353. cecal REVISION OF THE BRITISH WILLOWS. 457 Salix sphacelata, Sm., 386. splendens, Bray, 370. spuria, Willd., 430. Stephania, B. White, 424. stipularis, Sm., 414, 415. strepida, Forbes, 408. Stuartiana, Sm., 426. subdola, B. White, 354. subpurpurea-cinerea, Kern., 450. superata, B. White, 423. tenuifolia, Sm., 402. tenuior, Borr., 409. ephrocarpa, Wimm., 402, 406. tetrapla, Walker, 400. thymelsoides, Schleich., 411. rani, eng., 855, 358. iri P y 855, 358 triandra X alba, 354. triandra x fragilis, 348. triandra x viminalis, 355. undulata, Ehrh., 355. LINN. JOURN.— BOTANY, VOL. XXVIT. Salix vacciniifolia, Walker, 411. vaudensis, Forbes, 408. velutina, Schrad., 414, 418. venulosa, Sm., 410. viminalis, L., 413. viminalis X purpurea, 448, viminalis x repens, 391. viminalis x Caprea, 413. viminalis x triandra, 355. viminalis-purpurea, Wimm., 448. viminalis-repens, Lasch, 391. viridis, Fr., 303, 367, 371. vitellina, Z., 371. Wahlenbergii, And., 433. Waldsteiniana, Willd., 411. Wardiana, B. White, 402, 403. Weigeliana, Willd., 400. Woolgariana, Borr., 447. Synanpre, 447. Triandre, 347. Viminales, 413. 8 - ui Linn. Soc. Journ. Bor. Vor. XXV. Pu. 9. Buchanan Write - 70 70! qo p 65 oi — RT | | | | GO Ei = = 55 88] — —31——— 52. = = | ES 45 45 45 | = 1886 1890 1828 1830 1835 1838 1860 1873 1881 1884 1762 1804 B.W inv et del eantern Bros. hth Buchanan White. The squares 1, 2, 3, 4, 5 and 6 represent. true species. A (a combination of 1 and 2), B (of 3 and 4), & C (of 5 and 6) RR Nee represent hybrids of two species. D (a combination of ACly 2) 3 4 and, B o354)); and E (-Bc3 x4) x C(5x 6)) are hylnids of four Spectes . Er D (AX B) X E (Bx c) 16 a hybrid of sia spectes . Linn. Soc. Journ. Bor. Von. XXVI. Px. 10 | Hybrids of sta. species can also be preduced. in other WANS . as e.g. by A (1x 2) crossing with 3 to produce G(=1 273), and C (55x 6) crossing with 4 te produce H (4 75x 6) when E— ———3 4 — ——- Gx HF: or by COS XO) DCA aa 2 x Bexa). Iiybrids of live species may originate in various ways but one example 5 6 PLA My orguuere . pe erem] wlll suffice, C(556) x GABA AGEN =I (lo 242a 6). 7N see the. dotted. lines & squares, H \ Á N / EN / N, we N S ^ / x "d N / \ fy N / \ 4 Te = 7 7 ^ \ , D / I N 4 y | 2 / 1 3 4 / N | ^ Ly ^N 7 H H N + L N SC " TERM N Uv Pd UU N LN - 7 ed? N T P =~- N LM T7 s d ===- -— , pu ~- MN LA - T'si et) b -7 " E GH - -= NUT / — "gf We / Ps “>s... e ~~ / y -N CORO ; 2 V x KA ~~ Pd N Te >~ p NO Tse ~~ — \ =.. ~ / p \ =-~ 77. vole ` =-=. ~~ 7 S M. -~ - N ——-——L ~~ pO UT0—r-—--- N 77cm i i ] p------ EEN > L H 1 i | i "d H | 1 1 G i i | 016. H ` i H 1 i ( i 1 ! i i [ i i ' ! i i H I i i i i i i 1 “i bi 2 3 e i H H ' H | (11213 | 6 i EEN 5 6 1 ' 1 1 i \ < H i . i 1 H i i i | 1 t i ! l | ! i 1 ' i i ! i ' ! i i i LL BE DH H , ~ H H J— RT! B W inv. et dei. Buchanan White. Linn. Soc. Journ. Bor. Vou. XXVII. Pr. lH. LAPPONUM | | BN | ` PLEIANDILE. n DIANDRA. The Sections indicate : The larger circles indicate the three tribes; the smaller circles the species . LL net ascending above 1000 feet. . The lines connecting the species indicate that. Il, ascending fiou below 1000 feet to above. 2000 feet. these are known to form hybride. Ill, rarely if ever descending to 1000 feet. N, occasionally descending below 1000 feet. Mintern Bros. lith. B.W. inv. et. del ' . DIAGRAM OF THE HYBRIDS OF THE BRITISH SPECIES OF SALIX. acis REVISION OF THE BRITISH WILLOWS. 457 - sphacelata, Sm., 386. - splendens, Bray, 370. spuria, Willd., 430. Stephania, B. White, 494. stipularis, Sm., 414, 415. strepida, Forbes, 408. Stuartiana, Sm., 426, subdola, B. White, 354. subpurpurea-cinerea, Kern., 450. superata, B. White, 423, tenuifolia, Sm., 409. tenuior, Borr., 402. tephrocarpa, Wimm., 402, 406. tetrapla, Walker, 400. thymelæoides, Schleich., 411. Trevirani, Spreng., 355, 358. triandra, L., 347, triandra Xalba, 354. Salix vacciniifolia, Walker, 411. vaudensis, Forbes, 408. velutina, Schrad., 414, 418. venulosa, Sm., 410. viminalis, L., 413. viminalis X purpurea, 448. viminalis X repens, 391. viminalis X Capree, 413. viminalis X triandra, 355. viminalis-purpurea, Wimm., 448. viminalis-repens. Lasch, 391. viridis, Fr., 363, 367, 371. vitellina, Z., 371. Wahlenbergii, And., 433. Waldsteiniana, Willd., 411. Wardiana, B. White, 402, 403. Weigeliana, Willd., 400. Woolgariana, Borr., 447. triandra x fragilis, 348. triandra x viminalis, 355. undulata, Ehrh., 355. SYNANDRÆ, 447. Triandre, 347. Viminales, 413. Life-History of a Stipitate Freshwater Alga. By Grorce Masse. (Communicated by the Secretary.) [Read 5th December, 1889.] (Prate XII.) THE genus Dictyospherium was established by N ägeli * for the reception of a very remarkable freshwater alga—Dictyo- êpherium Ehrenbergianum, Nüg., found floating in ditches near Zurich. The generic diagnosis is accompanied by good figures and an excellent description of the leading morphological features of the plant, The following closely allied “ species ” are all that are known up to the present :— Dreryoseu xniv x EunENBERGIANUM, Nig. Hinz. Alg. p. 73, t. i. E; Rab. Alg. Eur. sect. iii. p. 47, f. 20 (dextra) ; Wolle, Alg. U. States, p. 186, pl. dei, ff. 29-30; Cooke, Brit. Alg. p. 20, pl. ix. f. 1. Dicerxospn ÆRIUM RENIFORME, Buln. in Hedw. ii. p. 22, t. 1. * Gatt. Einzel. Algen, p. 72, t. ii. E (1848). L INN. JOURN.— BOTANY, VOL. XXVII. 2K 458 MR. G. MASSEE ON THE LIFE-HISTORY f. 6; Rab. Alg. Eur. sect. iii. p. 47, f. 20 (sinistra); Wolle, Alg. U. States, p. 186, pl. clvi. f. 28; Cooke, Brit. Alg. p. 21, pl. ix. f. 2. Exsiec. Rab. Alg. no. 789. DicryospHarium Hrrcncocxirr, Wolle, Alg. U. States, p. 186, pl. clviii. f. 12. DicrxosPeuxniUM PULCHELLUM, Wood, Alg. N. Amer. p. 84, pl x. f. 4. Exsice. Wittrock et Nordstedt, Alg. Exs. no. 239. Two species are common to Europe and North America, whereas D. Hitchcockii and D. pulchellum have hitherto been observed in the last-named country only. The species are all very similar in general appearance, and under a low magnifying-power resemble minute hyaline spheres studded with brilliant green spots. During the past summer a Dictyospherium, which appears to agree with the spherical-celled variety of D. Ehrenbergianum figured by Cooke *, has been fairly common in the open-air tank at Kew, mixed with Spirogyra and Cladophora. Although the tank has been regularly searched for Algw for some years, nO trace of a Dictyospherium was seen before the present season. Last year the same tank was covered for some time with a dense orange-coloured scum consisting of the rare alga Spheroplea annulina, Ag.; whereas this year not a trace of this alga has been seen, although myriads of its oospores must have sunk into the mud. In eonnection with the last-named plant, it may be mentioned that just before its appearance the tank had been dry for some time for the purpose of cleaning and repairing. stocked two jars with Spheroplea obtained from the tank, one 0 which, owing to neglect, became dry during the summer, and remained in that condition for a week. The second has always contained water, and up to the present time no trace 9 Spheroplea has appeared in it; whereas in the jar that became dry, the Spheroplea appeared in quantity about ten days after it had been filled up with water. From the above statement it would seem that the oospores of Spheroplea benefit bya period of desiccation : as to whether this condition is absolutely necessary remains yet to be proved, and possibly similar conditions may to some extent influence the sudden appearance or disappearance of other forms of freshwater Alge. * Brit. Ale pl. ix. f. 1, c. OF A STIPITATE FRESHWATER ALGA. 459 In the earliest stage, the Dictyospherium caunot be distin- guished from a small specimen of Pleurococcus vulgaris, Meneg. The cell-wall is at first about lg thick, and consists of cellulose, becoming bluish-violet on the application of chlor-iodide of zinc. As the cell increases in size, but before it shows any signs of fission, the cell-wall inereases in thickness, and at the same time becomes mucilaginous, with indications of stratification, and in this condition the cellulose reaction is confined to the innermost firm layer, which corresponds to the original cellulose wall, the peripheral mucilaginous layer becoming brown with a solution of iodine, and behaving in many respects more like protoplasm than any modification of cellulose. The chlorophyll is of a very bright yellow-green, but the presence of chromatophores was not satis- factorily determined. The application of hematoxylin reveals the presence of a nucleus, especially after the specimen has been treated with absolute alcohol for a few minutes. When the chlorophyllose portion of the spherical cell attains a diameter of about 6u, it divides simultaneously into four equal parts by two septa developed in two planes at right angles to each other. The septa do not extend quite to the centre of the mother-cell ; consequently the four lobes into which it divides remain organi- cally united in the centre by thin portions about 2 diameter, and may be compared to four plums united by their stalks in the centre of the group. The mucilaginous portion of the mother- cell-wall does not divide along with the chlorophyllose portion, but continues to increase in quantity, and envelops the segments in à continuous hyaline stratum. After segmentation of the mother-cell the segments, at first more or less angular, become spherical, still remaining attached by their central stalk-like Portions, which continue to increase in size until they attain a length of about Lü ; these stalk-like portions are usually per fectly hyaline, with a very minute lumen, which under a high magnifying- Power is seen to contain granular fragments scattered here and there, which are sometimes coloured green. In old plants the stem-like connecting portions have the lumen completely obliterated. By the time the connecting Portions have reached the length of about 10p, the chlor ophyllose Portions of their enlarged apices have respectively assumed the *ppearance and size of the original mother-cell immediately fore segmentation ; each swollen apical segment of the cell is RoW surrounded by its own mucilaginous coat, the whole being 2x2 460 MR. G. MASSEE ON THE LIFE-HISTORY enclosed in the mucilage secreted by the unsegmented mother-eell. There is a total absence of septa, and the cell now resembles a pair of dumb-bells crossing in planes at right angles to each other, and enclosed in a hyaline sphere of mucilage. When the above-described stage of development has been reached, each of the four swollen apices becomes segmented into four portions; but this process differs from the method described as taking place in the mother-cell, in the segmentation into four lobes not being simultaneous. In the first instance, each swollen apical portion is divided by a cell-wall lying in the same plane as the long axis of its stem; the two segments diverge, each becoming stipitate, the stem-like portions starting from the apex of the original stem and forming a dichotomy. When the stems supporting the two segments have attained a length of about 8p, their swollen chlorophyllose apices again divide into two equal portions, by septa developed in a plane at right angles to those of the previous generation; the stems continue to increase in length and diverge as before, forming a second dichotomy. H each of the first four segments has divided into four portions, the cell now consists of sixteen globose chlorophyllose portions terminating an equal number of hyaline stem-like achlorophyllose threads radiating equally in all directions from a central starting- point, each of the four primary arms bifurcating, and each arm m turn giving origin to two others by a second bifurcation, the ultimate arms bearing at their free tips the chlorophyllose portions of the cell. The third time of division sometimes terminates the vegetative activity of the cell, but usually the segments divide in a similar manner four or five times, so that the cell consists respectively of thirty-two or sixty-four segments. As a rule all the swollen apices divide into two segments each, 1n rare instances three are produced, or no segmentation takes place, when the geometrical symmetry of the cell is disturbed and à departure from the typical sphere results. At tbe period of segmentation of any given generatiou of swollen tips, the mucila- gous cell-wall surrounding each loses its individuality, and mingles with the general sphere of 1aucilage surrounding the com plex cell. The vegetative phase terminates abruptly, and is a once followed by that connected with reproduction. The chloro- phyllose masses of protopiasm terminating the ultimate dicho- tomes become slightly contracted, at the same time the eh OF A STIPITATE FRESHWATER ALGA. 461 chlorophyll becomes aggregated on one side of the minute spheres of protoplasm, which externally exhibit rotatory movements, and finally escape from the cells as globose zoospores about An in diameter, furnished with a minute knob-like projection on the colourless side of the sphere, from which originates two exceed- ingly slender cilia, each about 12u in length. During the period occu pied in the differentiation of the proto- plasm, previously concerned with vegetative work bearing on the well-being of the individual, into zoospores, the mucilaginous wall, as also the innermost firm cellulose portion belonging to each one of the apical swollen portions, has lost its individuality and blended with the common sphere of mucus. The liberated zoospores are all of equal size, and after remaining active for about two hours become stationary, resorb their cilia and become surrounded by a wall of cellulose, during which change the chlorophyll becomes equally diffused throughout the cell, which !5 Now in a condition to recommence the vegetative phase by fission, It is certain that many succeeding generations are produced by the above method during the summer months ; as to Whether the ultimate zoospores become encysted, and in this Condition pass the winter months, I have unfortunately been unable to determine. Although the plant under consideration is technically uni- cellular, it nevertheless approaches very closely various generic types of freshwater alge that are probably, out of deference to historic prejudice, considered multicellular. he most prominent morphological feature presented by the species of Dictyospherium is the repeated division of the original spherical cell into portions radiating equally from a central point, and in the persistent concentration of the protoplasm at the tips of the repeatedly dichotomous radiating stem-like portions. Ip the last feature the present genus at first sight appears to be unique, but the difference is really but one of degree. The spores of numerous, in fact of most species of multicellular alge, both freshwater and marine, aud also of many unicellular forms, on germination, produce a distinctly stipitate plantlet, due to the Protoplasm remaining concentrated at the apex and leaving a colourless, thin, empty portion of the cell-wall behind in the form of a stem-like base, which is usually discoid at the point of attachment. Such forme differ from the initial stage of Dictyo- sphærium in the spore not at once dividing into four, but remaining 462 - MR. Q. MASSEE ON A FRESHWATER ALGA. intact, and thus forming one growing-point only instead of four. In all multicellular alge the stem-like base eventually disappears, but in the following genera, which are all freshwater forms, the attenuated and often elongated basal portion is permanent, hence the species are described by systematists as being “ stipitate ” :— Characium, A. Braun; Hydrianum, Rab. ; Sciadium, A. Braun (Sctadium is included as a section of the genus Ophiocytium by Rabenhorst); Peroniella, Gobi; Actidesmium, Reinsch ; Dictyo- spherium, Nüg.; Colastrum, Nüg.; Dimorphococcus, A. Braun ; Mischococcus, Nüg. ; Cosmocladium, Bréb. ; -Dactylococcus, Nig. ; Oocardium, Nag. The following genera form a transition from permanently stipitate to sessile forms:—W. Andrews, This species extends northward from the Straits 476 MR. JOHN BALL ON THE from Port Sta. Elena, Cape Fairweather, and Port Desire, but it has not been found on the west aide of the Andes. CRUCIFER A. 5. VESICARIA MENDOCINA, Phil. in Linnea, xxxii. 12,=V. andicola, Gill. MSS. in Herb. Kew.=V. montevidensis, Eichl. in Flor. Bras. xiii. 302, tab. lxvii. V. arctica, Hook. Bot. Misc. iii. 138, e£ Barn. in C. Gay, Fl. Chil. i. 161, non Richardson, Bot. App. Franklin’s Journey, 15, et Hook. Fl. Bor-Amer. t. i. “On the edge of table-land, sandy or stony ground. Native name Aleli.” —W. Andrews. There has been much confusion as to the name of this plant and as to its geographical distribution. It was first collected by Gillies on the lower slopes of the Andes near Men- doza, and sent to Sir William Hooker with the manuscript name V.andicola. lt was, however, supposed to be identical with the Vesicaria arctica of Richardson, and in the Hookerian herbarium, now ineorporated with the general collection at Kew, was laid on the same sheet with the North-American plant. Other spe- cimens from the same neighbourhood were sent by Cruckshanks, which are incorrectly marked Chili on the sheet in Kew Herb. (vide Hooker, loc. cit.). The same plant was collected by Tweedie in North Patagonia, aad by Sello in Uruguay. Eichler described Sello's plant under the name F. montevidensis in the part of the ‘ Flora Brasiliensis’? which appeared in December 1865 ; and in his ‘Symbolæ ad Floram Argentinam,' p. 16, Grisebach, recog- nizing the identity of Eichler's plant with that of Gillies, adopted the name V. montevidensis and rejected the name V. andicola, because the plant appears to grow only on the lower slopes of the Andes. In describing the plants collected by M. Claraz (Journ. Linn. Soc. xxi. 212) I called the plant V. andicola, Gill., because the objection stated by Grisebach is certainly invalid. But I did not advert to the fact that V. andicola is merely a manuscript name never published, and that apart from a single sheet in Kew Herbarium, there is no way of identifying it. Further than this, I was not then aware that early in 1864 Philippi, who had received specimens from the neighbourhood of Mendoza, published a correct description of it in the ‘ Linnea ' under the name V. mendocina, which name, under the law of priority of publication, it must now retain. The plaut has a wide range on the east side of the Andes, but FLORA OF PATAGONIA. 477 is probably most common in Patagonia, as it 18 inciuded in all the collections from that region, and has received a native Indian name. It is not properly included in the floras of Chili or Brazil. BIxacEz. 107. AZARA wricRoPHYLLA, Hook. f. Fl. Antarct. 944? Uncertain because the specimen has neither flower nor fruit, but closely resembling the foliage of that species, which is known only from Southern Chili and the island of Chiloe. A further reason for doubt is that so few species are common to the opposite coasts of the South- American continent. PoRTULACER. , 4. Porrunaca amaNDrELORA, Hook. Bot. Mag. tab. 2885. ‘Flowers in June and February ; abundant among the sandhills hear the coast of North Patagonia; flowers a brilliant magenta.” —W. Andrews. This appears not to be a common species, but has been collected at various spots in the territory extending from Mendoza to North Patagonia. Specimens collected by Balansa In Paraguay which I have seen in the Kew Herbarium apparently belong to an undescribed species allied to P. grandiflora. The flowers are said to be of a purple colour. 19. Porrunaca onmracea, Z. “Much esteemed by the natives as a cleanser of the blood, is eaten raw as a salad or boiled and dressed as spinach. Local name Verdu larga ; found chiefly hear rivers, aud also near the coast.” — W. Andrews. This 18 one of the most widely spread of cosmopolitan weeds, being found in almost all temperate and tropical parts of the earth, not only in continental regions and Australia, but also in Mada- 8ascar, the islands of the Indian Ocean, in most of the Pacific islands, and even in many remote oceanic islands of volcanic Origin, Tt may be conjectured that the agency of birds as well as that of man has been requisite to achieve the result. Another plant of the same natural order, Grahamia bracteata, Gill., was collected in Patagonia by Tweedie, and, more recently , 1n the valley of the Rio Negro by Niederlein. It is common in Some parts of Argentaria, but is apparently confined to the east side of the Andes, 478 MR. JOHN BALL ON THE MALYACEX. 102. SIDA sULPHUREA, A. Gray, Pl. Fendl. 23,=Malva sul- phurea, Gill. in Hook. Bot. Misc. iii. 149, Malvastrum sulphu- reum, Griseb. Symb. 43. Though not, as I believe, anywhere common, this species has a wide range on the east and west sides of the continent. 83. SPHÆRALCEA ——? nm Grows on the plains, flowers bright purple or magenta."—J7. Andrews. 64. SPHERALCEA ? “From the plains; flowers bright erimson."—. Andrews. I have no doubt that this and the last are forms of the same species, but I am not able with certainty to refer them to any of those described. The very numerous Malvacee of temperate South America, especially those belong- ing to the genera Malva, Spheralcea, Malvastrum, and Cristaria, urgently require careful examination by a botanist who can study the living forms; from incomplete herbarium specimens it 18 often difficult to fix even the generic position. These specimens are certainly nearly allied to Spheralcea bonariensis, Griseb. Pl. Lor. 45,— Malva bonariensis, Cav. I find it impossible to fix the limits between that species and S. cisplatina, St.-Hil. PI. Us. t. 52, and S. obtusiloba, G. Don, Gen. Syst. i. 465,— Malva obtusi- loba, Hook. in Bot. Mag. 2787. These are all very variable plants, especially the last, from which I think it is impossible to separate S. coquimbana, Phil., S. collina, Phil, and S. flexuosa, Gill MSS.in Herb. Kew. The specimens sent by Mr. Andrews have the lobes of the leaf more sharply cut than in any of the forms above enumeratéd, but I should conjecturally refer them all to a single very variable species which extends to both sides of the S. American continent, and which may best bear the name SPHERALCEA BONARIENSIS. 22. SPHERALCEA ? * Blossom of a brilliant scarlet or deep crimson, like a minute poppy; found in clay soil on low plains, especially near burrows of the Vizcacha."—W. Andrews. This is doubtless allied to 8. bonariensis, but cannot, I think, be referred to that species. The stem, which is woody at the base, is divided into numerous feeble trailing branches, the leaves are small pinnato-lobate with narrow segments, hoary on both faces. This is possibly an abnormal form growing in manured soil near the burrows ; the foliage much resembles that of a small specimen FLORA OF PATAGONIA. 479 collected by Captain King at Port Sta. Elena, to which the MSS. name Malva crispa, Hook. fil., is appended in Kew Herbarium. That plant, however, appears to be a Malvastrum, and the materials available for the study of this group are at present too seanty. In the botanical chapter of the ‘ Informe Oficial de la Expedicion al Rio Negro, published in 1881, Niederlein has described two plants from the valley of the Rio Negro, one of which he has named Malva patagonica, and the other is said be nearly allied to, if not the same as, Malva brevipes, Phil. TILIACER. 114. AnisTOTELIA Maqvi, L’ Hérit. This plant is common in Central and Southern Chili , having its northern limit (vide C. Gay, ‘Flora Chilena,’ i. p. 337) at the river Illapel, about 8. lat. 31? 30. Excepting this specimen, I have seen none from the eastern side of the continent save those collected by Tweedie in Uruguay. The genus belongs exclusively to the southern hemisphere, the other known five or six species being natives of New Zealand, Australia, or Tasmania. ZYGOPHYLLES. 27. LARREA DIVARICATA, Cav. “Shrub attaining a height of 12 or 15 feet, though generally less; always found on the high table-lands ; soil, sand and clay with pebbles ; disappears to- wards the west as soon as the mountain-ranges of Central Patagonia commence. It is of a highly resinous nature, and burns with a bright erackling flame; but the smoke possesses a strong acrid taste and odour, as does the foliage when bruised. he flowers, which appear in great profusion, are small golden stars, which in their turn give place to small round seed-globes Covered with a white cottony fur. The wood is very hard, and n many respects like box. This is one of the handsomest shru s of the uplands, presenting at one time a brilliant map of yelow at another (when in seed) of white, whilst the elegance o its Srowth always attracts attention. Local name Jarilla, but a 5o called Cordoba compass, as its growth is said invariably to be from south to north.” — W. Andrews. This species extends from Central Patagonia northward to many stations in Argentaria as far as Catamarca. Althou gh included in the ‘ Flora Chilena, I do not 480 MR. JOHN BALL ON THE believe that it has been found on the west side of the Andes. Specimens collected by Tweedie, and labelled South Brazil, are almost certainly from Uruguay. 27. LARREA CUNEIFOLIA, Cav. This has been sent by Mr. Andrews with the same number as the last species. It has nearly the same geographical range, but appears to be less common. 52. Larrea NITIDA, Cav. “Shrub growing on high dry ground, very resinous, burns with a bright flame, dense smoke, and acrid odour—small yellow blossoms which grow on the leaf.” —W. Andrews. Elsewhere Mr. Andrews says that this species is not found in the valley of the Rio Negro, but on the outer slopes of the Andes above the junction of the rivers Limay and Nenquen. . It extends northward to the outer slopes of the Andes both on the eastern and western sides to about S. lat. 30°. The secondary branches often bear four or five pairs of opposite leaves very regularly disposed, and Mr. Andrews apparently regarded them as forming a pinnate leaf. The flowers and fruit are borne on very short thick pedicels. GERANIACER. 94. GERANIUM PATAGONICUM, Hook. f. Fl. Antarct. 252. This appears to agree closely with authentic specimens from Southern Patagonia and Fuegia. Thosefrom theisland of Juan Fernandez, from Monte Video, and from Mendoza, which have been doubt- fully referred to this species, may perhaps be better ranked under G. intermedium, Bertero, unless they should all be regarded as forms of the same species. 108. Oxatis varprvrewsm, C. Gay? “Yellow blossom: shamrock-like leaves, tuberous root, and sharp acid flavour when chewed."— W. Andrews. Except by somewhat larger flower this does not seem different from the Chilian species. LEGUMINOS2Z. 32 and 96. @LYCYRRHIZA ASTRAGALINA, Gill. “ Plant of à dark green brilliant foliage; flowers of a pale purple ; always found on low ground and in the vicinity of water. The leaves exude a thick viscid humour when pressed, and communicate à FLORA OF PATAGONIA. 481 sweet flavour. The root yields a dark brown dye. The height Varies from 6 inches to 3 feet. Local name Urasu, or liquorice Plant.”—77. Andrews. "This plant seems to extend through a large part of temperate South America. It is nowhere common, but is less rare on the eastern than on the western side of the Andes. 6. ApzswrA murtcata, DC 5 Very imperfect fragments without fruit. 24 and 105. Viera GRAMINEA, Sm. in Rees’s Cyclop. xxxvii. “A small Vetch with blossoms of a deep brilliant blue with White at the throat. Found in valleys, soil clay and moist. Very widely distributed and abundant in mountain-ranges of Central Patagonia."— W. Andrews. The vetches of South America are very puzzling to the botanist. A considerable number of closely allied forms extend through the continent from the equator to the Straits of Magellan, if not also to Central America and Mexico, The typical form of V. graminea is, I believe, con- fined to the east side of the Andean chain, but closely allied Varieties or subspecies are found in Chili. 103. Latuyrus TOMENTOSUS, Lam. “Large blue blossom, amongst sand-hills near coast.” — W. Andrews. This appears to be a characteristic species of the east side of the continent, extending from Port Desire (Puerto Deseado) d northward to Uruguay. Specimens labelled “ Brasilia, collecte by Sellow, were probably collected in Uruguay. 14. LATHYRUS pubescens, Hook. 4 Arn., Bot. Bosch Fey LI. “White blossom, found along the banks of streams climbing among shrubs; especially abundant in the vicinity o Conesa on the Rio Negro."—W. Andrews. This species has a Wide range both on the eastern and western sides of the cone tinent, extending from the Chonos Archipelago northward to olivia and Peru, and from Patagonia to Catamarca. 80. CXSALPINIA Grugert, Bent. “ Shb ainoa a n M of 5 feet or alittle more. Flowers light yellow, from the = r of Which falls a group of bright scarlet stamens. The e is el tained in a pod similar to that of a pear, but 1s fab Foun tal tablelands in sandy soil.” — W. Andrews. Tu OTRU 482 MR. JOHN BALL ON THE plant appears to be not uncommon in some parts of Argentaria, and elsewhere to occur sparingly at rather wide intervals from Central Patagonia to Uruguay, and on the west coast extends somewhat north of the tropic of Capricorn. 25. HoFFMANNSEGGIA FALCARIA, Cav. “Small plant; height 4 to 6 inches, grows on clay soil in valleys of streams, though often at a considerable distance from water; flowers a dark orange, mottled with deep brown spots.”—W. Andrews. This is a small form of a species which extends to the northern Argentine States and also to the Chilian coast, generally growing to about 1 foot in height. 98. HoFFMANNSEGGIA TRIFOLIATA, Cav. “ Plant with small dark yellow blossoms, mottled in throat with spots of very dark purple growing in a spike; generally in neighbourhood of sea- coast or river.’ — W. Andrews. This species, rare in herbaria, 18 one of the few that appear to be peculiar to Patagonia. It was originally described by Cavanilles from specimens collected at Puerto Deseado. 29. CASSIA aPHYLLA, Cav. “Shrub leafless and low growing, height about 10 inches ; with yellow (dark gold) bell-like blossoms; always found on high table-lands in dry sandy soil.” — W. Andrews. This is apparently confined to the east side of the continent, ranging from Northern (or Central?) Patagonia to the Northern States of the Argentine Confederation. 43. PROSOPIS srRoMBULIFERA, Benth. in Hook. Journ. Bot. iv. p.352. “Dwarf shrub; height varying from 4 to 10 inches; root a very long straight tap descending perpendicularly. It 1s found throughout the plains of the Rio Negro valley, generally in clay ground. Blossom a reddish yellow ; fruit-pods of a bright yellow in closely twisted spirals, four or five growing together on a single stem. A decoction of these has a sweet, bitterish, and aromatic flavour, and is used both by natives and Indians as a sovereign specific for dysentery, diarrhoea, &c. Local name, Saco trapo."— W. Andrews. I am inclined to believe that the range of this species is confined to the eastern side of the Southern Andes from Catamarca to Northern Patagonia, but not approaching the Atlantic coast. There is a specimen sent by Mr. Reed from Camarrones in Chili in the Kew Herbarium; FLORA OF PATAGONIA. 483 but in C. Gay’s * Flora Chilena, Clos remarks that this species, which he wrongly supposes a native of Peru, is cultivated in gardens in Chili, probably for the sake of its medicinal properties. It is called in Chili Retorton. In the same work it is described as growing (in gardens ?) to a height of from 5 to 8 feet, very much exceeding the stature of the wild plant. The singular group of species which Bentham distinguished as a section of the genus Prosopis by the name Strombocarpa, and to which Asa Gray was disposed to give generic rank, seems to be limited to the eastern side of Extratropical South America (Argentaria and North Patagonia), and to the dry region of North America extending from Western Texas to Arizona and Northern Mexico, There are, indeed, two described species— P. reptans, Benth. in Hook. Journ. Bot. iv. p. 352, known only from the central provinces of Argentaria; and P. cinerascens, A. Gray, Pl. Wright. i. p. 61, from North Mexico and the ad- Joining region, which cau scarcely be separated. Bentham in Trans. Linn. Soc. xxx. p. 381, says of P. cinerascens, Fructiculus P. reptanti simillimus, nec distinguendus nisi pube evidentiore presertim in peduneulis, et interdum in floribus ipsis, et foliolis paullo maioribus et 14 lineam longis." Were it not for the wide difference in the habitat, Mr. Bentham would not have sepa- rated these plants, especially as in some specimens of each there 48 no apparent difference in the length of the leaflets. The legumes in Mr. Andrews's specimens are somewhat larger and thicker than in the other specimens of P. strombilifera which I have seen, but I should not distinguish it even as a variety. SAXIFRAGER. 66. EscanroxrA ——? “Shrub abundant on eastern slopes of the Andes."— W. Andrews. I have had no opportunity of observing the species of this genus on the eastern side of South merica; but I was much struck by the extreme variability of the forms encountered on the lower slopes of the Chilian Andes. it seemed to me that it would be very difficult, even when studied m the living state, to assign good diagnostic characters by which istinguish many of the species. I have ventured eg to express the belief that this is one instance, of which severa will occur to the student of the European Flora, in which the ‘Processes have not been completed by which, amongst a large 484 MR. JOHN BALL ON THE number of nearly allied vegetable forms, the more vigorous, or best adapted to surrounding conditions, are selected, while the less fit disappear. In such cases it becomes impossible to establish species with any approaeh to scientific clearness, and inducements are offered to those who find pleasure in coining new names. Professor Oliver is disposed to refer Mr. Andrews's specimen to Escallonia stricta, C. Gay, Fl. Chil., or to E. Philip- piana, Engl. I do not question his opinion, but I doubt whether these and some other so-called species should be separated from E. corymbosa, Pers. (Ruiz and Pav. Fl. Peruv. iii. p. 14, tab. 284). There is a specimen of this species apparently correctly named in the Kew Herbarium, collected by Matthews in the Peruvian Andes; but it would be necessary to examine a fuller series of specimens before expressing a positive opinion. Al- though the figure of E. alpina, Popp. et Endl. Nov. Gen. 1 tab. 13, differs considerably, I am inclined to suspect tbat this also should be united to the plant of the higher region of Peru. ONAGRARIER. 28. (CENorHERA STRICTA, Ledeb. “Plant with green fleshy foliage and light yellow flower. The blossom closes at evening, and only reopens in the morn after the sun has commenced to shine. Local name, Estrella de la vispera. The leaves of the flowers [petals] are boiled in oil, and this is then strained and thickened to an unguent which is highly esteemed in the eure of wounds. A plain decoction of the whole plant is also said to be highly beneficial for washing abrasions, wounds, &c. It 1s always found in sandy soil..—W. Andrews. The specimens collected by Mr. Andrews are whitish in the dried state from abundant but very short tomentum. 100. (sorgen), oporara, Jacg. “Plant growing to the height of 2 feet in sandy soil and near the sea-coast, with large yellow blossoms."— JV. Andrews. In both species the flowers of the living plants are described as yellow ; in drying, the petals of Œ. odorata have become rose-coloured. There can be no doubt that the native home of both the species here mentioned is in the American Continent, but their present distribution is cosmopolitan. Mainly, if not exclusively, through the agency of man they have been carried to nearly all FLORA OF PATAGONIA. . 485 the warm and temperate regions of the earth, and in many places they are so thoroughly established as to be commonly regarded as indigenous. LoaAsACEEX. 65. Loasa PrNNATIFIDA, Gill. ew Arn. in Edinb. Journ. Nat. d Geog. Sc. iii. p. 273 (1831) z« L. filicifolia, Popp. Fragm. 23. This is à mere fragment, but it agrees well with the original Specimen collected by Gillies. With regard to this as well as several other South-American Loasacew of which mature fruit is not found in European herbaria, I regard the generic position as qute uncertain. CUCURBITACES. 3 and 8. WILBRANDIA VILLOSA, Cogn. “Creeping plants; dark green foliage, small yellow blossom ; found on gravel and Pebble banks in vicinity of sea-coast."— W. Andrews. This species appears to have a. wider range than has previously been Supposed. All the specimens which I have seen come from the territory of Uruguay, and those quoted by Cogniaux "8. Brazil " are probably from the same region. Dr. Lorentz collected it in the province of Cordova; but it now appears that it extends far to the south. CacTAacEs. l. Cereus nEvcANTHUS, Pfeiff. Enum. diagn. Cact. p. 71= Echinocactus leucanthus, Gill. = Echinonyctanthus leucanthus, aire? “ Blossom large, white, with yellow eye, of a cactus. This stows up in a long round spike about 4 or 5 inches in circum- ference toa height of 2 or 3 feet ; from the top, or near it, springs the llower-stem."— JV, Andrews. Professor Oliver is disposed to refer the specimen to the same species which was collected near Mendoza by Gillies. It appears to me somewhat uncertain. 2. OPUNTIA GRATA, Phil.? “Blossoms yellow of a cactus, Šrows in large ovals about 6 inches long by 8 wide, the flowers Spring direct therefrom without any stem. Both this species and the last produce a purplish-red fruit of a sweetish taste. Side the appearance is that of a fig ; the fruit is most refresh- 486 l MR. JOHN BALL ON THE ing and nutritious. They are always found on high table-land ; soil generally sandy and stong — W. Andrews. The specimen seems to agree with one named by Philippi, and the species apparently inhabits both sides of the Andean chain. CALYCEREÆ. 7 and 38. Boorrs CRASSIFOLIA, A. Gray in Proc. Amer. Acad. v. 821; Hook. Ic. Pl. t. 1762, = Acicarpha crassifolia, Miers. * Plant with bright stiff leaves and a large yellow blossom, some- what resembling Helichrysum, in height from 1 to 2 feet ; always met with in sandy soil and most common amongst medanos, espe- cially those of the sea-coast, and occasionally found on the higher table-lands. One of the most common plants of Patagonia, extending south to the Straits of Magellan.” — W. Andrews. In the remarks which Professor Oliver has added to the de- scription in the ‘Icones’ above cited he points out that this species should be referred to Calycera rather than to Boopis, if the latter be maintained as a distinct genus. But be adds the expression of a doubt whether this should not be included in the original genus Calycera. Amongst the species now known there are many intermediate in structure between the typical Boopis anthe- moides and the true Calycera, as defined in Bentham and Hooker's ‘Genera Plantarum. There is nothing in the habit or the geographical distribution of the species referred to one or the other genus that tends to show that they should be separated. COMPOSITÆ. 4land 90. STEIRA SATUREIFOLIA, Sch.-Bip., var. ANGUSTIFOLIA, Baker. “ Plant resembling candytuft; height about 8 or 10 inches ; grows in sandy soil near the sea-coast, and generally in plains of Rio Negro valley; blossoms a mawkish pink.” — W. Andrews. This very variable species has a wide range on the eastern side of the continent, from Brazil to Northern Patagonia, but has not been found on the western side of the Cordillera. 93. GRINDELIA PULCHELLA, Dun. Mém. Mus. Par. v.51, tab. 6, =Q. diffusa, Gill. ex Hook. et Arn. Comp. Bot. Mag. ii. 45. “Plant with yellow flower, grows in sandy districts.” — W. Andrews. This species appears to be confined to the east side of the Andes. FLORA OF PATAGONIA. l 487 92. GRINDELIA SPECIOSA, Gill. ex Hook. et Arn. Comp. Bot. Mag. ii. 45, var, P “ Plant of large bushy habit, grows luxuri- antly amongst sand-hills near the coast ; large yellow flower.” — W. Andrews. Iam by no means sure that this belongs to the species found by Dr. Gillies in the interior of the continent near Mendoza. The scales of the involucre are far less unequal, the exterior being broader and the interior less broad than in the original plant. The leaves also appear to be different in texture and the venation much more conspicuous. G. speciosa has been found at many places on the E. side of the Andes and has not been found by the Chilian botanists: but one specimen from Gillies is labelled “ Andes of Chili.” 76. HAPLOPAPPUS CORONOPIFOLIUS, DC. Prod. v. 347,2 Diplo- Pappus coronopifolius, Less. This agrees perfectly with Chilian Specimens, Tt appears to be common in the provinces of Valdivia aud Concepcion ; but, excepting this, I have seen no specimens from the eastern side of S. America. 9, 99, and 116. Lxvucorsrs sericea, Baker in Mart. Fi. Bras. Vi. rrr. 7,= Aster sericeus, Less.,==Aplopappus sericeus, Hook. et Arn, Comp. Bot. Mag. i. 47. “ Plant with grey foliage, yellow blossom, developing subsequently into white down; common throughout plains, found also on high plains, but very stunted in growth.” — JV. Andrews. This is one of the few species that seem to have an equally wide range on both sides of the Andes in tem- Perate South America. The specimens marked 99 are nearly the same as Baker's variety eriophora, which is the same as Noticastrum ertophorum, Remy in C. Gay, Fl. Chil. iv. 20. 83. Baccmamrs waGELLANICA, Pers. “ Low-growing shr ub, growing amongst rocks and in moist parts of high ground. — W. Andrews. Hitherto I believe that this species 1s known ouly from the coasts of the Straits of Magellan. 42. Baconaris AnrEMISIOIDES, Hook. g Arn. “ pana Plant, silvery foliage, height about 1 foot; blossoms e, grey, afterwards turning to white down. Found in sandy sol , more especially by the immediate edge of the plateau ; n abundant between Puuta Rubra and Patagones; grows ; A rivers Negro and Chubat."— W. Andrews. With reference to the 488 MR. JOHN BALL ON THE above statement as to the colour of the pappus, I note that in most specimens that I have seen the pappus is distinctly rufous, especially in one collected by Tweedie at Bahia Blanca; in those collected by Mr. Williams, and in one from Monte Video, the colour of the dry pappus would be described as dirty white. The range of this species is considerable, but apparently confined to the eastern side of the continent. It extends from about the 44th to the 30th degree of S. latitude; but, although included in the ‘Flora Brasiliensis,’ has not, I believe, been found in Brazil. 48. BACCHARIS GLUTINOSA, Pers. “Shrub; height from 4 to - 6 feet; blossom white. Very general along the course of the Rio Negro; to the south is always met with in sheltered valleys and generally in the neighbourhood of water."—JV. Andrews . The distribution of this very variable species is very singular. On the west side of the continent it extends from the neighbourhood of Lima to Southern Chili through some 27 degrees of latitude. On the eastern side, where it is less common, it extends from Northern Patagonia to Uruguay and the Central Argentine provinces, but not to Brazil. Specimens that I have seen from Paraguay and Southern Brazil appear to belong to B. serrulata, Pers., which I believe to be a distinct species with longer, broader, and more numerous involucral bracts. 37. BACOHARIS NOTOSERGILLA, Griseb, “ Sbrubby plant, height from 1 to 24 feet; small yellowish-white blossoms ; has à very strong aromatic scent; found in sand and clay ground, always in valleys and low-lying parts; local name Escobilla. ls employed by the natives to make brushes similar to our besoms. Very widely distributed ; in places, as between Roca and Chin- chinal, forms almost the sole vegetation, and reaches right from the sea-coast to the foot of the Cordillera and southward to the Straits of Magellan."— JF. Andrews. This species has hitherto been known only from Montevideo and the Central Argentine DIV" vinces. It evidently has a much wider range, but until specimens have been received there must be some doubt as to its extension to the Magellanic region. Like the two preceding species, it 13 included in the * Flora Brasiliensis, although not, so far as We know, a native of Brazil. 6. BAccnanrs Gruss, A. Gray in Proc. Amer. Acad. v. 178, var.? “Small shrubby plant; dark green foliage; grows !? FLORA OF PATAGONIA. 489 sandy and salt ground; common in the lower valleys of the Chu- bat and on the sea-coast.”—W. Andrews. This specimen, named as above by Professor Oliver, seems to me very doubtful. It appears to be very nearly allied to B. valdiviana, Phil. in Linnga, rem, 138, 62. GNAPHALIUM SPICATUM, Lam. “Silvery-leaved plant growing in gravel."— W. Andrews. This species extends from Mexico through the west side of tropical America to the South- temperate zone, at least as far as Northern Patagonia. If I am night in believing that G. americanum, L., and G. purpureum, L., are forms of the same species, they should be united under the name G. americanum, and will be found to extend over an area of fully 80 degrees of latitude. 20. GAILLARDIA SCABIOSOIDES, Benth. et Hook. f. Gen. Pl. ii. 414-- Cercostylos scabiosoides, Arn. in DC. Prod. “ Plant bushy, about 15 inches high, grows in dense masses on low clay plains. e blossoms, which are without exterior petals, possess a strong odour resembling that of ripe apricots. These fresh, or powdered into a kind of snuff, are much used by the natives as & remedy for headache, Foliage of a dark green, somewhat white on the lower surface,” W. Andrews. In the « Symbolæ ad Floram Argentinam ' Grisebach correctly quotes the ‘Genera Plantarum’ for the name of this species, 15 the species is there named and distinctly referred to Gail- lardia. I have seen no specimens of the Brazilian plant with Undivided leaves (Guntheria megapotamica, Spreng. = Cercostylos brasiliensis of Lessing); but I have no doubt that Mr. Baker 48 been well advised in uniting the two as forms of a single "Peces, On further consideration I cannot, however, concur With him in reviving Sprengel’s specific name as that of the type of the collective species. The Brazilian form is rare, at least in Auropean herbaria, while G. scabiosoides appears to have a con- siderable range on the western side of temperate South America and to be comparatively common. (See Journ. Linn. Soc. Bot. XX. 223.) ** Small creeping plant, grows also on plains in neigh- aromatic odour." — 2x 63. SENECIO wrsmm, Hook. f. amongst rocks and on edges of springs, urhood of Port Desire; has a pleasant and LINN, JOURN.—BOTANY, VOL. XXVII. 490 MR. JOHN BALL ON THE W. Andrews. lbelieve that this species is confined to South Patagonia and the coasts of the Magellan Straits. E A specimen in the Kew Herbarium is labelled “ Maldonado, Tweedie ;” but it is allowable to suspect that the label has been misplaced. 69. SENECIO subuLatus, D. Don, ex Hook. et Arn. Journ. Bot. iii. 320, var.P “ Large bushy shrub, yellow blossoms, very luxuri- ant, found in sand-hills on sea-coast."— W. Andrews. This speci- men is in most respects intermediate between S. subulatus and S. linariefolius, Poepp. ex DC. Prod. The typical $. subulatus was first found by Dr. Gillies in the province of Mendoza, and it appears to be common in the pampas of the interior provinces. The same plant was distributed under the name Senecio mendoci- nus, Phil., but not published by Philippi. S. linariefolius is said to be common in the Chilian Cordillera. In the place above re- ferred to, Hooker and Arnott have united to S. subulatus a plant of which poor specimens were collected by Tweedie near Bahia Blanca. The specimens sent by Mr. Andrews are from a much larger and more vigorous plant, have larger heads and rather broader involucral scales, but must, I think, be united with S. subulatus, unless both be included as forms of Senecio linarte- folius, which is the older name. 110. SENECIO RANCONENSIS, Sch.- Bip. ad calc. Lechler Berberid., sine descriptione. “ Plant with yellow flowers, straggling growth, height about 2 feet ; grows under trees in thick woods on eastern slope of Andes."— JF. Andrews. This agrees very well with Lechler's specimens from the Andes of Ranco, so that the species seems to inhabit the opposite slopes of the same portion of the great range. In his ‘Catalogue of the Chilian Flora’ Prof. Friedrich Philippi has amended the specific name to the more correct form S. rancoensis. 31. Srngcro HvarrATA, Bertero. “ Plant growing with à large upright stalk ; height 4 or 5 feet ; leaves large, green and fleshy, the bigger leaves being at the lower part of the stem; flower- stalks shoot out from the upper part, bearing groups of white blossoms with a yellow ẹye, in appearance similar to but larger than the common horse-daisy ; always found in low moist grou? and ground the edges of lakes."— W. Andrews. This conspicuous FLORA OF PATAGONIA. 491 species extends to both sides of temperate South America, but is more common on the Pacific slope. 84. Murisra ILICIFOLIA, Cav. Te. v. 68, tab. 493. “Creeper; large bright orange blossom with dark centre. Slopes of the Andes and pre-Cordillerine districts."— W. Andrews. 85. Apparently the same plant; a single head. “Pink and carmine blossoms, also occasionally white." — W. Andrews. This is not different from M. truncata, D. Don, and some forms are not easily distinguished from M. spinosa, DC. It is singular that this species is us ually quoted “M. ilicifolia, Hook." Sir W. Hooker published a description of the plant in the ‘ Botanical Miscellany,’ and in so doing referred to the description and figure Previously given by Cavanilles. This species extends to both slopes of the Andes. 49. Dette ARGENTEA, D. Don, ex Hook. et Arn. Comp. Bot. Mag. i. 108. « Bushy plant, very general in all northern parts of Patagonia ; grows in sand, especially amongst the medanos (sand- hills) near the coast. Foliage silvery white, flowers pink. Local name Majai: height about 2 feet.” —W. Andrews. This orna- mental species seems to be rather common in North Patagonia and extends thence to the Province of Mendoza; but it is very doubtful whether it bas been found wild in Chil. 36. Cuugurraca ERINACEA, D. Don? “Shrub: prickly foli- age, bright yellow blossoms in great abundance: very widely dis- tributed, being found in all classes of soil, as well on high table- lands as in valleys, on sea-coast and in the interior. Branches of this boiled yield a yellow bitter fluid which the Indians take as a febrifuge, In Chubat the settlers have used it instead of hops in making a sort of beer.” —W. Andrews. T do not feel sure as to the identity of this species, nor do I unreservedly accept the statement that the same species 18 found "ety herein Patagonia. There are numerous nearly allied forms of ulicine Chuquiragas which seem to require careful study by 8 botanist possessing ample materials. 17. Basket, CHRYSANTHA, Mik. “Blossoms of bush found n * sheltered valley near Lake Lajara in the Andes. in ` n ? feet.” — yy Andrews. This is the most interesting object 1n the Collection made by Mr. Andrews. This very 23 species has M 492 MR. JOHN BALL ON THE hitherto been found in Brazil, chiefly on the Corcavado and other erninences near Rio de Janeiro, and in the Province of San Paulo. The new habitat must be at least 1600 miles distant. 34. BRACHYOLADOS LYCIOIDES, Gill. ex D. Don in Phil. Mag. March 1832; Hook. et Arn. Comp. Bot. Mag. i. 106. " Shrub, height from 2 to 4 feet; hard dark green foliage; blossom of & deep gold-colour. Found chiefly on high table-lands on clay or stony soil, occasionally in valleys. Common on theSenger, Chubat, Deseado and Rio Negro."— W. Andrews. This species was origin- ally found by Gillies towards the base of the Andes of Mendoza, and was soon after collected by Tweedie in North Patagonia. It is enumerated in the * Flora Chilena; iii. 812, and is said to grow in the Cordillera of the province of Santiago. I have seen no Chilian specimens, unless a single specimen in the Kew Herbarium, labelled “Bridges, Concepcion,” belongs to the same species. This has the stem and branches erect, not spreading, nearly quite glabrous; the involucral scales nearly glabrous and the medial nerve much less prominent ; and finally the achenes much larger. In the Patagonian plant the oblong achenes have five obtuse rounded ribs; in the Concepcion plant the fruit is not quite mature, but the achenes appear to have much more angular ribs. In this plant the branches bearing the flowering heads are often beset with projecting knobs which are abortive branches, and frequently bear a single minute leaf about an eighth of an inch in length. This is especially noticeable on the specimen from Concepcion above mentioned. | 70 and 88. TRICHOCLINE INCANA, Cass. “ Plant of small and compact growth, with yellow flowers. Generally found on open plains and in sandy soil; sometimes on clay.”—W. Andrews. This species has a wide range iu eastern South America, extending to the Peruvian Andes. it has not been found in Chili, and it 18 very doubtful whether any species of this genus extends to the western slope of the Andes, (See * Flora Chilena, iii. 288.) 106. Leucerta RUNCINATA, Gill. ex D. Don, in Phil. Mag. 1882 (sub Leuchæria)= Chabræa rosea, DC. Prod. “ Plant with thistle- shaped leaves; flowers a pale mauve; grows in thick woods on eastern slope of the Cordillera; attainsa height of about 8 feet." —W. Andrews. This is an Andine species which has descended for some distance on both slopes of the great range. FLORA OF PATAGONIA. 493 101. Levceria ACHILLEIFOLIA, Hook. et Arn. Comp. Bot. Mag. ii. 48 (sub Leuchzria) = Chabrea multifida, DC. Prod. “Plant growing in sandy soil; flowers a pale mauve."— W. Andrews. The original specimens from which this species was described were collected by Darwin at Port Desire. DeCandolle founded his Chabrea multifida on a specimen from the same place collected by Née. It may be doubted whether several of the described species of this group should not be united. The specific name, as well as the generic, proposed by Hooker and Arnott is entitled to precedence, 78. PEREZIA nEOURVATA, Less. = Homoianthus echinulatus, Cass. =Perdicium recurvatum, Vahl. “Small low-growing plant; blossom bri ght blue; found amon gst mountains in sandy and stony districts.” — Hr Andrews. This species has been supposed to be confined to the Magellanic region. It may be inferred from the above note that it extends some way northward towards the base of the Andes. 94. NASSAUVIA ROSULATA— Acanthophyllum rosulatum, Hook. d Ara. Comp. Bot. Mag. ii. 48. “Low creeping shrub growing in clumps on high gravelly clay or rocky plateaux. Not found till South of the Senger."— W. Andrews. Although not in flower or fruit, there can be no doubt as to the identity of this singular Species, as to which I have given some particulars in the ‘J ournal of the Linnean Society,’ Bot. xxi. p. 225. VACCINIACER. l 86. PERNETTYA ? “Shrub of low compact growth; berries of a wax-like appearance shading from pink to white; found near streams or near trees in the neighbourhood of the Andes. —W. Andrews. The single small specimen resembles most P. mucro- nata, Gaudich., but cannot, I think, be referred to that species. PLUMBAGINEX. l 82. STATICE BRASILIENSIS, Boiss. “ Plant growing in moist low ground, and always indicates the presence of salitre (salt an Nitrate of Soda) in the soil. The flower is lightish blue verging to pale pink or white. The root is large and long like that o the dock. Boiled in water it gives a decoction of a dark brown colour and salt taste. It is an invaluable remedy for all interna bruises, tumours, &c. Local name ‘ Wykeru.’ »—W. Andrews. 494 MR. JOHN BALL ON THE It 1s impossible not to suspect that the species of this group found on both coasts of the western hemisphere should all be referred to the original Statice Limonium of Linneus. Along with S. caroliniana, Walt., inhabiting the eastern coast of North America from Newfoundland to South Carolina and Texas, Boissier has described in DeCandolle's ‘ Prodromus ’ S. californica from the Pacifie coast, and S. brasiliensis, which extends ou the east coast of S. America from the neighbourhood of Rio Janeiro to Southern Patagonia. To these Philippi has added a fourth species, S. chilensis, Linnea, xxxiii. 220, apparently a rare plant which he had received from only one locality. Of the forms here enumerated the first is the most different from S. Limonium, but Boissier's S. californica is, in many respects, intermediate. The characters assigned to these Species are scarcely sufficiently stable, and may most of them be paralleled among the undoubted varieties of S. Limonium from different parts of Europe and the Mediterranean region. Boissier is disposed to attach importance to the presence or absence of hairs upon the calyx-tube, but I find this character to be very variable. In some specimens of S. californica the calyx is quite glabrous. Ihave not seen authentic specimens of S. chilensis, unless a specimen from Coquimbo and another from Valparaiso in the Kew Herbarium should be referred to it. Both specimens are quite immature with the flowering branches undeveloped, but are not distinguishable from S. Limonium. The branches of the panicle in 8. brasiliensis are described as scabrous; in Mr. Andrews’s specimens the branches in the dried state are rough, but certainly not scabrous. LoGANIACER. 118. BUDDLETA GLOBOSA, Lam., var. foliis subtus albo-tomen- tosis. “Local name Retamilla. Grows in abundance on slopes of the pre-Cordillera. I have plants of this raised from seed I sent to England in 1882." — W. Andrews. 'Thisis, I believe, quite new to the eastern side of South America. All the other specimens I have seen are from Chili; and in these the lower surface of the leaf is covered with a reddish-brown tomentum. ASCLEPIADER. 35. PHILIBERTIA GirnLresit, Hook. et -Arn.=Sarcostemm Gilliesii, Decne. in DO. Prod.=8, incanum, Decne. l. és, Do- FLORA OF PATAGONIA, 495 manum, Decne. l. c, “Creeper, found on high ground and amongst shrubs of table-lands; flowers a round bell-shape, of a dirty white, mottled in the interior with very small spots of purple ; foliage a bright green.’—W. Andrews. This is not a creeper, but a climbing plant with weak branches that twine round any support, or, failing such, round each other. It seems to extend over a wide area on the eastern side of South America. 26. AscLEPIAS LINIFOLIA, Decne. “Plant bushy; fleshy foliage of a bright green; flowers a greenish white. The stem, when broken, yields an immense quantity of an acrid milk-like fluid ; height about 10 inches ; grows in a clump in clay grounds, and generally indicates the presence of salitre."— W. Andrews. I am unable to distinguish the limits of the species nearly allied to A. mellodora, St.-Hil., which are widely spread on the eastern side of temperate South America. CoNVOLVULACER. 111. Cresga CRETICA, L. “ Small plant; silver-grey foliage and whitish flowers; found in the island of St. Blas ; soil sandy."— W. Andrews. In these specimens of this cosmopolitan species the fruit is larger than usual. See the remark in Benth. et Hook. f, ‘ Genera Plantarum,’ ii. p. 881. SOLANACEAE. 15. NrenEMBERGIA ricipa, Miers. “Plant found in abund- ance on dry sandy or gravelly plains. These specimens were collected about 7 leagues to the S.W. of the Bahia St. Blas. Flowers a pure white, with purplish veins, and an eyeof the same colour.” — W, Andrews. "Though not without doubt, I am disposed to agree with Grisebach (* Symb. ad flor. Argent.’ p. 242) in uniting N. filicaulis, Lin dl., with N. rigida. A specimen labelled by him N. filicaulis undoubtedly belongs to N. rigida. This species seems to extend from Paraguay and the Province of Mendoza to North Patagonia. Some caution is necessary in endeavouring to ascertain the geographical distribution of South-American plants. This species and N. linifolia, Miers, are included in the ‘ Flora Chilena, though not, as I believe, found further west than the neighbourhood of Mendoza. In DeCandolle’s ' Prodromus Decaisne has described a plant as from South Brazil, adding that the specimens came from the banks of the Rio Negro of Patagonia. 496 MR. JOHN BALL ON THE 74. FABIANA IMBRICATA, Ruiz et Pav.? “ Shrub, attaining a height of 10 or 12 feet; grows in sandy soil, and most luxuriant in neighbourhood of Nahuel Huapi, a lake on eastern side of the Andes."— W. Andrews. The specimen is without flower or fruit, but it is scarcely possible to mistake this characteristic Chilian species. I have seen no other specimens from the eastern side of the Andes. 23. NICOTIANA ANGUSTIFOLIA, Ruiz et Pav.=Petunia acumi- nata, R. Grah.= Nicotiana acuminata, Hook. Bot. Mag. t. 2919. “ Plant growing to a height of about 4 feet; long slender-throated flower, opening with a cup; flower yellow; soil clay ; found on low- lying plains.”— W. Andrews. There is, I think, no doubt as to the identity of the species described by Graham with that of Ruiz and Pavon. As Remy remarks, the figure in the ‘ Flora Peru- viana' is ill-executed and misleading. The species is found on both sides of the Andean chain, but is more common in Chili. : 89 and 112. SOLANUM ELEAGNIFOLIUM, Cav. "Dwarf shrub, with glaucous foliage and yellow spines ; flower of a pale lilac, somewhat resembling that of the potato; fruit a small round yellow berry, with seeds iuside; locally called Manzana del Diablo. Height from 6 to 15 inches, generally found in clay ground in the valleys of Rio Negro, Chubat, &¢.”—W. Andrews. This species has a wide range in subtropical and temperate South America, being equally common on both sides of the Andes. ScROPHULARINER. 81. CarcronamrA PLANTAGINEA, Sm.?=C. biflora, Lam.? “ Broad-leaved Calceolaria, found at foot of the mountains. The leaves grow in a compact mass close to the ground. The flowers shoot up on stalks, much as the Cowslip."— W. Andrews. The specimen consists of a single leaf and two detached flowering scapes each bearing three flowers on very long filiform peduncles. I do not feel assured of the identity of the species. O. plantaginea is found in many places in the Chilian Andes, and extends south- ward to West Patagonia, but does not approach the Atlantic coast. 44. LINARIA CANADENSIS, Dum.-Cours. " Plant found on margin of Rio Negro, generally on gravel-banks ; flowers of a FLORA OF PATAGONIA. 497 pale purple."— W. Andrews. Widely spread through the tem- perate regions of North and South America. 109. Mrwvrvs nuTEUs, Z. “ Yellow blossom, grows in moist localities.” — W. Andrews. This polymorphous plant, of which several varieties have been described as distinct species, is also common to the northern and southern temperate regions of America. 53. SCUTELLARIA NUMMULARIÆFOLIA, Hook. Jf- "Small plant growing in moist valleys; flower deep purple with white centre.” — W, Andrews. This species has hitherto been found only on the shores of the Straits of Magellan. VERBENACER. 22. VERBENA BONARIENSIS, L. “ Plant with tall upright stem to a height of from 4 to 6 feet; very small deep-purple blossoms in clusters at point of long stems. Decoctions of this are much used to bathe bruises. Foliage ofa dark green, and rough to the touch. Grows in clayey ground, especially near streams, lakes, &c."— YF. Andrews. This species appears to be a genuine cosmo- Polite, extending to the warm temperate regions of the earth, though doubtless in many stations (for instance, the Canary Islands) it has been introduced by human agency. 47. VERBENA TENERA, Spreng.=V. pulchella, Sweet V. Ber- teri, Schau. “Plant with growth similar to Verbena; flowers white, also lilac; generally found in clayey soil."— W. Andrews. I have no doubt that V. Berterii is one of the forms of Sprengel’s F. fenera, a species very widely spread in South America. A Tevision of the species of this group, with the assistance of more abundant materials than those which were at Schauer's disposal, would, I think, lead to a great reduction in the number of admitted species. 68. VERBENA THYMOIDES, Phil., sec. schedam in Herb. Kev. ? “Tuft of a small, low and compact-growing plant, having an aromatic flower like thyme; flowers white, sometimes tinged with purple. Grows chiefly on high dry soil and amongst rocks. —W. Andrews. A plant which appears identical with this was collected at Port Desire by Darwin. It can scarcely be separated from a Chilian plant béaring the label above quoted ; but I have 498 MR. JOHN BALL ON THE seen no published description. Very near to this, and perhaps an extreme form of the same species, is V. bryoides, Phil., FI. Atac. 40. PLANTAGINER. 95. PLANTAGO BRIDGESII, Decne., var. ANGUSTIFOLIA, Oliver, MSS. “Plant found on clay or sandy plain; minute purplish flower ?"—JW. Andrews. CHENOPODIACEX. 40. ATRIPLEX cristata, Mog., var.? “ Creeping plant of a vivid green foliage, common throughout Patagonia; emits a strong pungent odour when bruised."— W. Andrews. I much doubt whether 4. Pamparum, Griseb., which appears to be com- mon in the Argentine pampas, can be separated from A. cristata. This has a wide range in America, extending to the Southern United States. PROTEACES. 87. EMBoTHRIUM CoccINEUM, Forst., var.? * Found on slopes of Andes in sheltered vales; of free high growth; blossoms red, shading in parts to yellow.”—W. Andrews. Professor Oliver is doubtfully disposed to refer this to the .E. lanceolatum, Ruiz et Pay. The specimens are very incomplete, and only a few flowers are developed. When better known, this may be found to belong to an undescribed species. E. coccineum is common in the Magel- lanic region and the southern extremity of the continent. E. lanceolatum has hitherto been found only in Chili. LORANTHACES. 110. LORANTHUS ? an L. tetrandrus, Ruiz et Pav.? “Shrub found growing in woods of the slopes of the Andes; blossoms dark red."— W. Andrews. This very imperfect specimen is doubtless very near to the Chilian Z. tetrandrus, which is a common parasite on the western slopes of the Andes, but has not hitherto been found elsewhere. SANTALACEZE. 97. QuiNCHAMALIUM MAJUS, Brongn, “Plant growing in sandy soil; many small heads of yellow in a close clump."— W. FLORA OF PATAGONIA. 499 Andrews. As the authors of the ‘ Genera Plantarum’ remark, IF is very difficult to define the limits of species in this genus, aud therefore to discuss their geographical distribution. Quin- chamalium majus certainly extends to both sides of the Andean chain. 76. ARJONA PATAGONICA, Hombr. & Jacquinot. “ Plant of low growth, on sandy and stony ground in mountain-districts; pale Purple flowers in tufts.” — 77. Andrews. This is a characteristic species of the Magellanic region, which extends some way north in Patagonia, but apparently does not reach Southern Chili. EUPHORBIACER. 18. Eupyorsra prosrrata, Ait.? “ Creeping” [should be prostrate] “ plant, growing most luxuriantly in stony clay ground ; small yellow flowers.” — W. Andrews. This cosmopolitan species is Closely allied to the Old-World E. Chamesyce. I was disposed to refer the present plant to E. Engelmanni, Boiss. in DC. Prod. XV. II. p. 42, which has been found abundantly in Argentaria and Tuguay ; but that species is distinguished by fimbriate stipules, Which Mr. Andrews's specimens do not seem to possess. CUPULIFERE. , SÜÜand 117. Faavs ost, Mirbel. “ Leaves of tree of large 8760; rough bark: eastern slopes of the Andes."— W. Andrews. Although leaves only have been sent, there can scarcely be any mistake as to the species. It extends to both slopes of the des, 67. Faaus ? “ Leaves of a tree of large size picked on the shores of Lake Lajara at beginning of autumn, at that time of an intense red colour."— W^. Andrews. Professor Oliver is disposed to refer these leaves to F. procera, Poepp. et Endl. ; but they appear to me quite uncertain. AMARYLLIDEE. 10. ZzemvmawTHES ANDERSONI, Herb. “ Crocus-like flower, of a bright magenta colour, rising from the bulb; foliage very Tare. Found all over the plains of Patagonia, slopes of the pre- Cordillera, and even up among the perpetual snows of the true Cordillera. I have gathered specimens on the Quattropillan at a 500 ON THE FLORA OF PATAGONTA. height of 3440 metres. Occasionally two flowers are found ona single stem, one slightly below the other. Local name Margarita rosada.” — W. Andrews. 11. ZEPAYRANTHES ?, an sp. nov.? “Small yellow Crocus-like flower; bulbous root; found in gravel, sand, and clay; height about 4 inches. The flower is generally without foliage; when present, this consists of two, rarely more, thin grass-like leaves.” — W. Andrews. Mr. Baker pronounces this to be near Z. filifolia, Herb. inedit., and probably a new species. 104. HirPEASTRUM ——? “Grows abundantly on grassy plains; flower white, varying to purplish carmine."— JF. Andrews. A single flowering scape, without root or leaves. GRAMINES. 89. MELICA RIGIDA, Cav. This species seems to be confined to the east side of temperate South America. 45. Bromus vxroLorpxs, Kunth. ‘ Grass picked near river in vicinity of Guardia Pringles."-—)/. Andrews. This grass 18 common throughout the greater part of South America. Friricrs. 51 and 60. AsPIDIUM CORIACEUM, Sw. Both specimens were collected near Port Desire. A cosmopolitan species, but chiefly limited to the southern hemisphere. eee UR STUDIES IN VEGETABLE BIOLOGY. 501 STUDIES IN VEGETABLE Brotoey.—VI. An Investigation into the True Nature of Callus:— The Vegetable-Mafrow and Ballia callitricha, Ag. By Spencer Le M. Moonz, F.L.S. [Read 6th March, 1890.] (Puare XIV.) HISTORICAL INTRODUCTION. Discoverxp rather more than half a century ago by Th. Hartig *, and studied, among others, by such renowned botanists as Von Mohl, N ágeli, Hanstein, De Bary, Janczewski, and Stras- burger, there is yet some doubt, not indeed concerning the mor- Phology of sieve-tubes, for that has been the subject of many a fine memoir, but doubt as to the nature of the substance found so constantly upon the sieve-plates, and which, since the appear- ance of Hanstein’s + monograph, has been known as callus. And not only this, but two distinctly opposite views are yet held as to the origin of this substance, one being that it is formed by simple swelling-up of the sieve meshwork itself, with subsequent coales- cence of the swollen portions into a continuous mass; while, according to other authors, callus owes its origin to the proto- plasm, or, to speak more correctly, the slime of the sieve-tubes, Which deposits it upon the sieve-plates. To follow this latter subject, first: the older writers, such as Von Mohl t and Nägeli $, may be claimed as advocates of the former of these views; for they seem to consider the sieve itself as the “middle layer” of the intertubal wall, the two callus-plates forming the wall proper. he next writer on the list, De Bary ||, with his usual caution, refuses to dogmatize on the subject, remarking that it requires further investigation. "Three years after the publicatiou of the ‘ Vergleichende Anatomie’ of De Bary, his pupil Wilhelm d ." ‘Fortschritte der Forstwissenschaft 1837. I have not seen this pub- lication, t ‘Die Milehsaftgefásse und die vorwandten Organe der Rinde,’ 1864. t Bot. Zeitung, 1855, $ Sitzb. der k. bayer. Akad. der Wiss. zu München, 1861. | Vergl. Anat., Engl. transl. p. 176. ; NA ‘Beiträge zur Kenntniss des Siebrührenapparats dicoty ler Pflanzen, 80. 502 MR. S. LE M. MOORE'S STUDIES published an elaborate monograph on the sieve-tubes of Vitis vini- fera, Cucurbita Pepo, and Lagenaria vulgaris ; in this Wilhelm expresses his conviction that callus is altered cell-wall; and this view was acquiesced in by Janczewski * shortly afterwards. In 1882 we find the opposite idea broached by Russow t, who takes his stand upon the ground of the difficulty of reconciling the disappearance of callus in spring-time with clearing of the sieve- plates, that is, their return to the primitive condition—the difficulty of reconciling this with the origin of callus by swelling- up of the wall of the sieve, which ought, if the older view be correct, to show at least some signs of waste in consequence. It is true that Janczewski looked upon the process as one of hydra- tion—the callus swelling up in autumn, a time when the watery reserves in the plant tend to increase, and condensing and con- tracting by loss of water on the return of spring. But this is obviously unsatisfactory ; for it may be remarked that were this idea correct, we ought to find callus diminishing in quantity pro- portionally to the withdrawal of water from the plant, and callus ought not to make its reappearance until quite late in the year. Russow ends a strong argument by remarking that all difficulties will vanish if it be admitted that callus consists of contents of the sieve-tube deposited upon the sieve-plate. Strasburger f came to Russow's support in 1884; and in the same and also following year Gardiner $ declared himself upon the same side: his example was followed by Fischer in 1886 Il- On the other hand, we find F. W. Oliver f, in 1887, reverting to the older view, his conclusions being founded on careful study of the remarkable deposits of callus in Macrocystis and Nereo- cystis. Finally, Lecomte ** has expressed himself in the same sense. * Mém. de la Soc. des Sc. Nat. et Math. de Cherbourg, t. xxxii, 1881. (Also an extended extract of this in Ann. Sc. Nat. Bot. 6 sér. t. xiv.) t Sitzb. der Dorpat. naturforsch. Gesellsch. 1882. Translated in Ann. Sc. Nat. Bot. 6 sér. t. xiv. This was the idea of Hanstein ( Milehsaftgefásse). 1 ‘Bot. Practicum,’ ed. 1 (1884). $ Proc. Camb. Phil. Soc. 1884, p. 102 ; ibid, 1885, p. 230. ` || Ber. d. deutsch. bot. Gesellsch. 1885, and Ber. üb. d. Verhandl. d. k. sächs. Gesell. d. Wiss. zu Leipzig, 1886; abstract in Bot. Zeitung, 1886, and in Journ- B. Micros. Soc. (same year). €| Ann. of Bot. vol. i. ** Bull. Soc. Bot. de France, t. xxxv. (1888), and Ann. des Sc. Nat. 7 sér. t. 10 (1889). ae ee nt POUR IN VEGETABLE BIOLOGY. 508 À third view has recently been propounded by Rendle *, who, by a thorough examination of the Onion, believes the callus in this plant to owe its existence partly to the wall, partly to the protoplasm. It is necessary to give this brief historical résumé, because, as vill hereafter appear, some of the facts brought forward in the present memoir are, it is believed, of such a nature as to put an end, once and for all, to the conflict in opinion which has prevailed upon the origin of callus, so far at least asregards the two types here dealt with. With reference to the other matter, viz. the nature of callus, we find De Bary + noting the yellow colour it takes with iodine dissolved in potassium iodide, its swelling in hydric sulphate till its outline is completely lost, and its swelling up in caustic potash. Russow ¢ appears to have been the first to hazard an Opinion as to the chemical constitution of callus: having hit upon the method of staining it with aniline-blue, and also with that mixture of iodine and Schulze’s solution now known as Russow’s reagent, he remarks § that it is proteids, and especially nuclein, that callus most closely resembles. Unfortunately, Just at this time Szyszylowicz || discovered the brilliant colour taken by callus in corallin-soda—a fact which set observers upon a wrong tack altogether. It was assumed that because starch and muci- lage both take up corallin-soda, because the protoplasm of the Bleve-tubes uudergoes, after the disappearance of the nucleus, a change into slim y matter, and because this change is accompanied by solution of the starch, for these reasons it was assumed that callus is a sort of starchy mucilage. This idea is, at least as far as the facts hereafter brought forward go, a wholly incorrect "De: and I will now proceed to defend this statement. * Ann. of Bot. vol. iii. t Vergl. Anat. Engl. Trans. p. 175. 1 Sitzb. der Dorpat. naturforsch. Gesellsch. 1881. $ L. c. (1882 memoir), , . | Vide Janczewski, 2. c. (Note Additionelle). Also a Polish memoir by Ber szylowicz, which I have not seen, but which is abstracted in Bot. Central- blatt, Band xii. p. 138, 504 MR. 8. LE M. MOORE’S STUDIES Prorerp REACTIONS OF THE CALLUS OF THE VEGETABLE-MARROW. The material used in this research was taken from a Vegetable- Marrow plant at the fag end of the season; by this means a plen- tiful supply of callus was ensured, irrespective of any which might be formed by cutting of the stem in pieces and by its pre- servation in aleohol, which, as Fischer * has shown, are methods very apt to increase the amount of *Sehlauchkopf " and of callus in the sieve-tubes. Fischert has described “ ectocyclie,” “ ento- cyclic,” and “ commissural ” sieve-tubes in Cucurbita, in addition to those of the vascular bundles; my observations refer to the latter alone. The three chief tests for proteids are the xanthoproteic reaction, depending upon the deep yellow colour produced on boiling with nitric acid and after-addition of ammonia, the red coloration imparted by Millon’s reagent upon warming, and the blue or blue-pink by copper sulphate and caustic potash or soda, the latter becoming deeper on boiling. Unless a substance gives all these reactions, it cannot be considered to be a true proteid ; if all three are yielded by it in the typical way, the presumption is that protein is present. . There is no difficulty whatever in getting the xanthoproteic reaction, if only the precaution be taken of adding an excess of ammonia. The method pursued throughout the research where boiling was necessary was the following :—Sections were mounted in the fluid appropriate to each case; a cover-slip was then put over them, and heat was applied by means of a spirit- lamp. If much of the fluid evaporated before the desired effect was obtained, more was run in, the process being continued as long as was deemed essential. The advantage of this method is that the operator always has his material well in hand, and 80 can prevent over-boiling—a very important matter, as will after- wards appear. In testing with nitric acid and ammonia, then, 1t is necessary that the latter be in excess: the effect of boiling 18 usually to cause some swelling of the callus, but not much; it frequently happens that the sieve-piercing connecting threads ar? brought beautifully into view by this means. The colour, which is a well-pronounced yellow, is markedly deeper than that, * Ber. d. deutsch. bot. Gesellsch. 1885. ted Y ‘Unters. üb. d. Siebrohrensyst. d. Cucurbitaceen, Berlin, 1884. Abstr iu Journ, R. Micros. Soc, 1885, p. 477. I have not seen Fischer's memoir. ese sss sn toti Re IN VEGETABLE BIOLOGY. 505 assumed by the slime, as will be noticed on a reference to fig. 1 of Plate XIV. There is one point in which this reaction is more satisfactory than the others, inasmuch as it readily admits of the Preservation of the tissue treated by it; the specimen shown to-night has been mounted in glycerine for several months, and lts colours are still as brilliant as ever. Millon’s reagent acts with Vegetable-Marrow callus as clearly as does the just-mentioned test. To obtain it, however, care must be taken to avoid over-boiling; indeed, the best effect is got at temperatures just below boiling; but it boiling be con- tinued even for a minute, the red colour is liable to disappear (Plate XIV. fig. 2), But it is more difficult to prove that, in respect of the third test, Vegetable-Marrow callus comes within the proteid group. If previous observers have tried this reaction, their failure is scarcely blameworthy on the score of remissness ; for who would dream of upwards of an hour elapsing before the effect is brought about? And yet this is indeed sometimes the fact. You mount the sections in copper sulphate, and, after a little time, run in caustic potash, but in almost all cases without any result for at least ten minutes. Then perhaps a callous mass at the edge of the prepa- ration will become very pale pink or bluish; this will gradually deepen into the colour of Plate XIV. fig. 8, or will become lavender ; after a few minutes two or three more will follow suit, until, finally, all are coloured. But it may happen that one hour, °r sometimes even half as long agaiu, must elapse before the slightest change is caused in any of the callous masses. Why this lapse of time should occur is difficult to understand: possibly the method is at fault: and there may be some way of rapidly getting the reaction; but if there is, it still remains to be discovered. All the above-mentioned reactions were given in a very clear manner by the slime of the sieve-tubes. The only difference observed was that the respective colours were less pronounced, 'n consequence, apparently, of the less dense aggregation of the Slime as compared with the callus. REACTIONS OF THE WALL OF THE SIEVE-TUBES. (a) Proteid Reaction.—Throughout the research this matter Was carefully attended to: the reason for this will be explained LINN. JOURN.—BOTANY, VOL. XXVII. 2N 506 MR. S. LE M. MOORE'S STUDIES hereafter, when treating of Ballia callitricha*. The only proteid reaction yielded by the sieve-tube tissue was the xanthoproteic ; but I could never find a trace of reddening with Millon’s reagent, nor did copper sulphate and caustic potash or sugar and sulphuric acid tinge the walls to the slightest extent. (b) Cellulose Reaction.—On the other hand, the walls of the sieve-tubes and cambiform and companion cells give the purple colour with Schulze’s solution, by which cellulose is known. The slime aud eallus are coloured brown by this reagent. Action OF a Pepronizina FLUID UPON VEGETABLE- Marrow CALLUS. Vegetable-Marrow callus thus gives, in the most distinct way, the chief reactions employed in the detection of proteids. And here arises a question of some interest, namely, whether this callus-proteid resembles the main body of proteids in being acted on by pepsin and typsin in an acid and alkaline medium respectively with formation of peptone; or whether, like lardacein, it is able to withstand the action of proteolytic ferments. In order to answer this question, experiments were arranged in the following way. Into the first of four vessels was placed some artificial gastric juice prepared according to Pharmacopeia directions, and some radial or tangential sections of the phloém of the Vegetable- Marrow. The second vessel contained some gastric juice with finely-chopped meat; this was to serve as a test of the good- ness of the pepsin. Sections of phloém with some 0'2 per cent. solution of hydrochloric acid were in the third vessel ; the object of this was to ascertain whether the acid alone would have any action upon the callus; and, if any, what action. The fourth vessel contained distilled water alone. All four vessels were then exposed to temperatures varying between 352-39? C. From the second vessel the peptone reaction was obtained after a little time, thus showing the genuineness of the pepsin. Sections from the first vessel were examined from time to time; and it s008 became evident that dissolution was iu progress. So much was this the case, that at the end of 81 hours a considerable part of the callus had already disappeared; while after 16 hours no trace of it remained upon the sieve-plates. The sieves were now * Vide infra, p. 514. IN VEGETABLE BIOLOGY. 507 perfectly open, just as at the beginning of the vegetative period, and they gave the purple colour with Schulze's solution. It has not been thought necessary to draw the stages of this process; but anyone who wishes to see figures of what occurs will be able to satisfy his curiosity by referring to tab. 9 of Janczewski’s memoir *, where dissolution of the callus of Aristolochia Sipho at the beginning of the season is beautifully figured. Fig. 10 on . Plate XIV. shows the sieves eleared of their eallus: in this ease the action was allowed to continue for 30 hours, by which time neither the callus in the acid solution, nor that in the water, had been in any way affected. In this second point, therefore, we see that the callus of the Vegetable-Marrow agrees with the great body of proteids. We have already found Russow drawing a parallel between callus and proteids, on aecount of the behaviour of the former to iodine and to aniline-blue ; and there can be no doubt but that Russow was upon the right clue, and that the starchy-mucilage view must, at least so far as relates to the Vegetable-Marrow, be definitively abandoned +. , It is to be presumed that peptone is formed as the result of the digestion of callus ; but this I was unable to ascertain, apparently because of the extremely small amount of callus dealt with. It Would seem that to get the peptone reaction with clearness, a much larger quantity of callus than I had must be employed. ORIGIN AND CLASSIFICATION OF VEGETABLE-MARROW CALLUS. The chief objections which have been urged against the swelling- up theory of callus-formation are, that the edge of the sieve is always visible, although plenty of callus may be present ; that in the progress of obliteration the connecting channels, at first rela- tively broad, become thinner and thinner, the sieve-plate mean- while not itself increasing in size $: and in connection with this, Russow § noticed that the rod-like stris visible in callous masses * Mém. Soc. Sc. Nat. Math. Cherbourg, xxxii. (1881). * Gardiner also at first supported Russow (Phil. Trans. 1883, and Pres Camb. Phil. Soc, 1884, p. 101); but he afterwards changed sides (Proc. Camb. Phil. Soo, 1885, p. 230). Hillhouse (Midland Naturalist, 1884, p. 122, note) Rotices some resemblance, “probably, however, merely casual,” between the reactions of callus and those of nuclein. . , . 1 Fischer, 1886 Memoir. Figs. 30-32 of this memoir are very instructive. $ Sitzb, Dorpat. Naturf. Gesellsch. 1882, 2x2 508 MR. S. LE M. MOORE'8 STUDIES upon treatment with iodine surround threads of slime, and, according to Fischer, this is due to conversion of the slime into eallus, the process taking place from without inwards. In addi- tion to this, it has always been a stumbling-block in the way of adherents to the other theory, that the sieve-plates of many perennial plants are cleared in the spring. This one cannot understand as happening if the callus be merely trausformed sieve-plate; and, besides this, as already mentioned, Fischer finds an increase in the amount of “Schleimkopf” and callus from mechanical injury and from preservation in alcohol, which is expli- cable on the deposition theory alone *. However, although, in my opinion, the latter theory has always been the stronger, the battle between the two views has necessarily been a drawn one, because continuous observation of a sieve-plate was impossible. But in the new method of employing à peptonizing fluid we have the means of keeping under observation one and the same sieve-plate, which we can see gradually clearing, and, finally, quite free from callus. We can then examine a sieve-plate which but a few hours previously was obliterated with callus, and find it in exactly the same condition as before callus made its appearance upon it. And when we remember that the callus is, at least in the Vegetable-Marrow, undoubtedly a proteid, and, as such, cannot possibly be formed from cellulose, it is submitted that the question is definitively settled, so far as the Vegetable-Marrow is con- cerned. With what class of proteids should Vegetable-Marrow callus be assorted? To answer this question, it is necessary to know its behaviour to neutral saline solutions: those employed in this research were sodium chloride and potassium nitrate, in each case in 1 per cent., 10 per cent., and saturated solutions. ln no case did the callus betray auy tendency to dissolve. From its insolubility in water and in neutral saline solutions, as well asin dilute acids and alkalies, while it dissolves in strong hydrie sul- phate, and, to some extent at least, in caustic potash, Vegetable- Marrow callus recalls, and may perhaps be classified with, the Coagulated Proteids. * Here, too, one may mention Janczewski's observation, that sometimes the sieve-tubes of old rhizomes of Phragmites communis have their sieves open 1n winter. This is supposed by Janezewski to be due to the tubes no longer con- taining protoplasm. But why should disappearance of the protoplasm prevent formation of callus, if mere swelling-up of the wall is involved ? ———— À IN VEGETABLE BIOLOGY. 509 A Suorr Account or Ballia callitricha, Ag.* The thallus of Ballia callitricha consists of a main axis com- posed of a single series of cells, each of which bears on either side a secondary axis, of which the basal cell is formed near the upper end of the parent-cell: these secondary branches lie in two series, the individual members of both series lying in the same plane as their predecessors. Each secondary branch, in its turn, branches in the same way as its parent; and so with branches of later order; and the whole of the ramifications lie in one plane: the plant therefore consists of a series of systems of superposed binary verticils all with their median plane coinciding. This arrangement is afterwards complicated by cortication of the axis, which commences by downward growth of the basal cell of each branch. The septa between contiguous cells of the same order of rami- fication are saddle-shaped, the side of the saddle being visible in the ordinary view. This accounts for the peculiar appearance of each septum, from which one might conclude that the upper cell intrudes upon the lower, whereas, as Archer f first showed, pre- cisely the reverse is the case. The saddle is shown in Plate XIV. fig. 4; in the other figures it has been thought sufficient to draw its upper part alone. At this upper (central) part are situated the remarkable bodies known since the date of Archer's paper as “stoppers ” : these are highly refractive as seen in dried specimens moistened, and more or less hemispherical in shape with a flat base by which each is applied to a sort of bed of cell-wall. There is, however, no adhesion between stopper aud wall, for the former can easily be detached from the latter, under which circumstances, if the means taken to separate them have not been too violent, the stopper may be seen to be hanging by plasmatic threads to the bed. By Appropriate means (Sehulze's solution, xanthoproteic test, Millon’s reagent, hydric sulphate, Schulze’s macerating fluid) à single plasma-bridge or several threads of plasma can be seen running from stopper to stopper ; these threads do not appear to traverse the stopper itself but to run out to its edge, and by this means neighbouring protoplasts are put in communication. * Vide Archer in Trans. Linn. Soc., 2nd series, Bot. vol. i. t Loc. cit. 510 MR. 8. LE M. MOORE’S STUDIES The phycoerythrin of the cell is, of course, soluble in water, but in dried specimens I find many of the cells without any trace of colouring-matter. The cell-protoplasm of dried speci- mens moistened is very often most beautifully reticulate; this reticulation is only exceptionally seen in young cells, but it is common in older ones ; indeed it is rare to find one of these cells without some trace at least of a reticulum. The meshwork may be tolerably even, or it may consist of large meshes with narrow bars in one part of a cell, and of smaller ones with thick bars in other parts, or both forms may be mixed. It frequently happens that the protoplasm is aggregated near one end or near both ends of the cell*; starch-grains of varying size are present in the meshwork. I have been unable to find a single nucleus in these cells—hsematoxylin and acetic methyl-green stain in every cell several small bodies in the way characteristic of nuclei, and hence it is to be presumed that we have realy to do with syncytia; but it is scarcely safe to dogmatize upon the minute structure of the cells of a plant which has undergone desic- cation. REACTIONS OF THE STOPPERS. The stoppers react in the following ways :— Acetic Acid has no effect upon them whatever, neither has strong Nitric nor Hydrochloric Acid. Strong Hydric Sulphate causes slight swelling, but so slight as to be often almost inappreciable. Hydrochloric Acid gives no red coloration to the stoppers, neither before nor after they have been soaked in a solution of phloroglucin, Carmine preparations : of these three were used—picrocarmine, Which imparted a yellow colour, and Beale’s and Thiersch’s carmine, which had no effect. Caustic Potash when cold may cause very slight swelling, but sometimes none at all; when not, dissolution is involved, but without exudation of yellow drops. . Fuchsin stains the stoppers brilliantly and permanently ; this I find also to be the case with ordinary callus. * This was noticed in my memoir on Protoplasmie Continuity. (This Journal, vol. xxi.) IN VEGETABLE BIOLOGY. 511 With Watery Eosin the stoppers are coloured a bright and permanent pink ; Hematoxylin also stains them bright red. Jodine dissolved in potassium iodide gives a rich brown colour, similar to that given by Russow’s reagent. Schulze’s solution also colours brown and without swelling the stoppers. Corallin-Soda stains beautifully, but, as is the case with ordi- nary callus, the colour soon fades, and after a time disappears entirely. Sands’s Picric Blue.—This is the aniline blue referred to in my memoir on Protoplasmic Continuity. It is an admirable reagent for callus, which it stains a peculiar blue, a blue much lighter than that taken by protoplasm. Curiously enough, the stoppers scarcely ever show any sign of blueing. A slight local stain may occasionally be seen upon them, but this is probably due to the protoplasm of the cell; from the reagent in question, the stoppers take up picric acid alone, becoming yellow in conse- quence. On running over the list of reactions, the resemblance in most essential points between Ballia stoppers and Vegetable-Marrow callus will at once be realized, and this is further emphasized by resemblance in optical properties, both substances being iso- tropic. The chief differences are :— 1. Insolubility of the stoppers in strong hydric sulphate. 2. The rich brown with iodine. 3. Refusal to take up aniline blue. m Of these, the first is obviously immaterial in its bearing upon the proteid nature of the stoppers ; the second being a distinctive proteid reaction, is an argument in favour of the proteid ped Refusal to take up aniline blue is certainly a peculiarity, but think that when all the evidence in favour of the proteid P has been marshalled, this solitary reaction will be regarded as o no effect upon the question. ACTION op A PeProxiziNG FLUID UPON THE STOPPERS. To determine this point, experimen already detailed when treating of the Veg , H Mona th markable difference in respect of gastric digestion tl for stoppers and Vegetable-Marrow callus soon came di 5 earin instead of gradually dwindling away and finally disappoa ` the stoppers gave no sign of the presence of a digesti , s were made in the way etable-Marrow. A re- 512 MR. S. LE M. MOORE'S STUDIES and this even after the lapse of 30 hours, long before which time the callus had all been dissolved away. However, it was deemed advisable to continue the experiment for some time longer; but inasmuch as, after 60 hours, the action of the fluid was still resisted, we ean hardly err in coming to the conclusion that the stoppers are not attacked by a peptonizing fluid. Plate XIV. fig. 7 shows a pair of stoppers still unaffected after 30 hours' action of the fluid. Prorsrn REACTIONS OF THE STOPPERS. We have thus seen the stoppers resisting gastric digestion, hence particular care is necessary in drawing conclusions as to their proteid nature. Our chief guide in this matter will be similarity in reactions between the stoppers and the protoplasm of the cell *. It will be necessary also to attend in some detail to the reactions of the cell-wall. Before entering on the task, however, I may say that the stoppers, both here and in other Floride in which they have been studied, have generally been considered as being of proteid, if not of protoplasmic nature t. Xanthoproteic reaction.—'This is very well given indeed; as with the Vegetable-Marrow, it is necessary to have a greal excess of ammonia if the reaction is to come off properly (fig- 5). The same colour, only somewhat paler, is given by the cell- protoplasm. Preparations treated in this way preserve their colours perfectly in glycerine. l Millon’s reagent produces a fine red-crimson colour in the stoppers, but if boiling be continued above a second or two, the colour flies ; it is best got at temperatures a little below boiling. The cell-protoplasm almost always takes a mueh paler colour, but where the contents are densely aggregated, as so often occurs in the neighbourhood of the stoppers, I have sometimes seen a colour almost, if not quite, as decp as that assumed by the stoppers. Comparison of figs. 2 and 4 will show the different ways in which Vegetable-Marrow callus and the stoppers react to * This point is insisted on by Fischer in his paper on proteid reactions of the cell-wall (Ber. d. deutsch. bot. Gesellsch. 1887, p. 426). + Thus Gardiner (Proc. Camb. Phil. Soc. 1884, p. 106) says, à propos of the stoppers, “the portion of the pit-protoplasm next the pit-closing membrane is usually well differentiated.” ebe" ta IN VEGETABLE BIOLOGY. 513 Millon’s fluid. One might perhaps, considering this colour- difference and considering too the way in which the bright hue of the stoppers flies if boiling be continued—a peculiarity, however, which is shared by the cell-protoplasm—one might, perhaps, regard the crimson of the stoppers with suspicion, the normal colour for proteids being more or less of a brick-red. But it must be remembered that there is much variation in this Tespect even among typical proteids; thus Krasser* gives the following instances of this fact :— Albumen takes a flesh-red. Fibrin » carmine. Legumin ` flesh-red to brown-red. Vitellin — ,, brick-red. The real reason for the bright colour I believe to be that the stoppers consist of densely aggregated matter and so would necessarily be highly coloured; and this idea is sustained by an already mentioned fact, viz. the deeper hue taken by the more densely aggregated parts of the protoplasm. Copper Sulphate and Caustic Potash also act very well; the effect is usually got within a few minutes. The colour is usually a pale dirty pink, but sometimes it is pale blue; the former is not a good colour for a proteid, it is admitted, but precisely the same colour is taken by the cell-protoplasm. Raspail’s reaction (Sugar and Sulphuric Acid) succeeds very fairly—the colour is a pale brown-pink ; the protoplasm behaves similarly. I have not tried Krasser's t Alloxan reaction, since Klebs ł has shown it to have no value in the detection of proteids. Neither has Adam Kiewicz's test been resorted to (violet colour with glacial acetic acid and strong hydric sulphate), for this is said to give too pale a colour to be appreciable under the microscope. There is just a chance, however, in view of the brilliant colour taken by the stoppers with the xanthoproteic and Millon 8 reactions, that AdamKiewicz's test might succeed with the stoppers. * Sitzb. d. Wien. Akad. Band 94. t Krasser, l c. 1 Bot. Zeitung. 1887. CN, Sette E E ONN 514 MR. S. LE M. MOORE'S STUDIES REACTIONS or THE CELL-WALL OF Ballia. AXanthoproteic reaction.—This is given very fairly, but not nearly so well as is the case with the protoplasm, and à fortior? with the stoppers. Wo other proteid reaction is given by the wall. Hydrochloric Acid neither alone nor after phloroglucin gives a red colour to the wall. With Schulze’s solution local blueing is caused, but the effect is not general even after twenty-four hours. On the other hand, boiling for a minute or two in caustic soda causes the whole wall to take a deep blue-black colour. This points to the presence of cutin in the cell-wall, and the action of Fuchsin mentioned below confirms this idea. Sands’s Picric Blue stains yellow the outer part of the wall of large cells—the inner part taking a dark blue, and the whole, or nearly the whole, of the intercellular septum is so stained. Watery Eosin stains the wall a brilliant permanent pink. Corallin-Soda also stains, but temporarily. Fuchsin colours the outer part of the wall brilliantly and permanently. Strong Hydric Sulphate causes the inner layers to swell up greatly. Hematoxylin stains the wall the characteristic bluish colour. We may perhaps suspect from these reactions that there is present in the cell-wall some substance or substances, either of proteid nature or a decomposition product of proteids; but the entire failure of Millon’s and the Copper Sulphate and Caustic Potash tests, tells very much against this. SOME OTHER Reactions OF WALL, CELL-PROTOPLASM, AND STOPPERS. It is scarcely to be doubted that, as Archer* maintained, and as I have myself also done my best to show t, there is no pit-closing membrane in Ballia callitricha. This fact comes out clearly when testing for Millon's reaction, the proteid contents of the pit taking the brilliant colour of the stoppers and no break being observable in the connecting column (Plate XIV. fig. 4). The * Trans. Linn. Soc. 2nd Series, i. . t Journ, Linn. Soc., Bot. xxi., also Wright, Trans. R. Irish Acad. 1879; Hick, Journ. of Bot. 1884; Massee, Journ. R. Micros. Soc. 1884, &c. For the contrary view seo Schmitz (Sitzungsb. d kénigl. Akad. d. Wise, zu Berlin, 1883) and Gardiner (Proc. Camb. Phil. Soc. 1884, p. 104). Siecle uae ics if IN YEGETABLE BIOLOGY. 515 n of boiling for a minute with Millor's fuid is shown at - ig. 8; here the colour has been discharged, but the protoplasmic reticulum (drawn in the lower cell alone) is still visible. The limiting layer of protoplasm comes out now very clearly ; it seems to be reflected over the stoppers *, but perhaps the appearance giving rise to this idea is really due to denser consistence at the boundary of the stoppers; at any rate the limiting layer also extends over the bed whereupon the stoppers rest, The latter now have a granular look, in sharp contrast with the uniformly white and apparently homogeneous swollen cell- wall; the granules are evidently precipitated by the reagent, and the protoplasm itself has much the same granular character. The Pit is here very well seen with its broad debouching points, and the connecting protoplasm also comes out very clearly. Another excellent way of studying the relation between stoppers and cell-wall is to use Schulze’s macerating fluid. With this, the cell-contents contract after the stoppers have become invisible. Sometimes only one ball of protoplasm is found in the cell, but more often there are several, the sphe- roidal form greatly predominating; each ball is surrounded by its own delicate membrane, and usually the pit is left quite clear of plasma (Plate XIV. fig. 9,4). In other cases, and especially if the acti on has not been too violent, the connecting protoplasm is still Visible (fig. 9, a), or there may be two or three threads, this being 4 consequence of coagulation of the original thread. When, as n fig. Ho, plasma-balls remain in position at either end of the Pit, there is no trace of the stoppers; the protoplasm of these balls is still reticulate. But the fact of most interest is that on now running in iodine, the plasma-balls at once take a fine brown colour—colour which is quite as pronounced in these balls Info which the stoppers have melted as it is elsewhere; the cell-walls remain uncoloured by the iodine. ORIGIN AND CLASSIFICATION OF THE SUBSTANCE OF Ballia STOPPERS. On the ground therefore :— o 1. Of the general appearance of the stoppers, which is so different from that of the cell-wall ; 2. Of their distinctive proteid reactions, those of the wall being but feebly shown, if at all ; * Thus Archer describes the stoppers as lying outside the primordial utricle ns. Linn. Soc. 2nd Series, i. p. 214). 516 MR. S. LE M. MOORE'S STUDIES 8. Of their behaviour resembling in almost every point that of the cell-protoplasm, and differing in almost every point from that of the wall ; 4. Of the evidence of differentiation as shown by boiling with Millon's fluid—this not being shared by the wall ; 9. Of the behaviour of the stoppers with Schulze's macera- ting fluid, viz. dissolution into adjacent plasma-balls, wbieh subsequently stain brown with iodine while the wall remains colourless ; it seems impossible to avoid the conclusion that, although they resist gastric digestion, the stoppers consist of proteid. It also seems impossible to believe them to be formed by swelling-up of the wall, for why should such marked proteid reactions, and such different behaviour generally, be shown by the swollen portions alone? We have also seen the pit cleared of its plasma and its mouths freed from stoppers; and the appearance of the wall, in this case, is such as to put an emphatic veto on the swelling-up idea. Study of development of the stoppers, so far as it is practicable from dried specimens, teaches the same lesson. Let us look at a young cell—the youngest in which the stoppers are visible: we see here a tiny sphere of substance, looking exactly like the protoplasm and quite unlike the wall, placed at either end of a fine and apparently continuous pit, of which the lumen is occupied by a thread of protoplasm. These spheres we can see increasing gradually in size as we pass away from the growing region, and in no case can the least participation of the wall in this increase be detected. At length after cortication has set in, the stoppers become of relatively enormous size, and frequently have bands of slimy substance attached to them ; this substance reminds one of the * Schlauchkopf ” of sieve-tubes, but it may perhaps be true protoplasm. The point to notice here is, that not even in these old cells, the stoppers of which give precisely the same reactions as the younger ones, can any swelling-up of the intercellular wall be detected. I see no way therefore of escaping the conclusion that, like the callus of the Vegetable-Marrow, the stoppers owe their existence to deposition at the pit’s mouth of proteid matter derived from the proto- plasm. Should the substance of the stoppers—which, like Vegetable- Marrow callus, is inscluble in neutral saline solutions—be included 2 Gr RICE IN VEGETABLE BIOLOGY. 517 among true proteids, or relegated to the miscellaneous group of bodies allied to proteids? Bearing in mind that all three proteid reactions are given, I think it would be better to include 1t among true proteids, where, on account of its resistance to gastric digestion, it will take its place alongside of lardacein. Of the bodies allied to proteids with which the substance might Possibly be compared, Mucin gives only two of the proteid reactions—it swells up strongly in water, and is readily soluble in alkalis. Chondrin is soluble in hot water, in alkalies, and ammonia. Gelatine swells up in cold and dissolves in hot water and in weak acids and alkalies, and it has, moreover, extremely feeble proteid reactions. Elastin, with feeble proteid reactions, is soluble in Hydric Sulphate and Hydric Nitrate, whereas Nuclein ives but an indistinct xanthoproteic reaction, and none at all with Millon’s fluid. These differences are all of them of such importance, that there can hardly be a doubt of the desirability of bud classifying the substance of Ballia stoppers with these odies. Some General Considerations. Sachs’s * theory that it is in the sieve-tubes that proteids are Constructed, has not found much favour. The general view now is that adult sieve-tubes are merely distributors of proteids, not manufacturers of them, for they soon become enucleate t. The Notion hinted at by Sachs 1 seems to be a good one; the callus according to this is interposed as a mechanical hindrance to the Passage of the contents of the tubes; in support of this may be cited Briosi’s § observation that starch-grains sometimes actually stick in the sieve-plates, as a proof that some mechanical hin- Tance is necessary. I cannot help thinking that Ballia Stoppers must function in the same way. What is certain is that the size of the pit increases with the growth of the septum, 80 the older a cell is, the wider will be the pit separating 1t from adjacent cells. The lower cells of the main and secondary branches ultimately become of great relative size, and must * Vorlesungen üb. Pflanzen- Physiologie, Vorlesung. no. Xx. m bad bp P. 325). That sieve-tubes are organs for transport of protei Wis 1861; H. von Mohl, Bot. Zeitung, 1855; Nágeli, Sitzb. bayer. Acad. Wiss. 3 Hanstein, Mitchsaftgefásse; Sachs, Flora, 1863, &c. ` t Fischer, Unters. üb. d. Siebrohrensyst. d. Cucurbitaceen. 1 Vorlesungen, no. xxii. (Engl. transl. p. 361). § Bot, Zeitung, 1873. 518 MR. S. LE M. MOORE’S STUDIES contain a considerable amount of proteid, though whether proteid is still being manufactured in them I cannot say; but in the oldest cells which eame under my observation were to be seen numerous darkly-staining bodies which are most likely nuclei. The branches are arranged in binary verticils, and the basal cell of the branch being considerably younger than the cell of the braneh of higher order supporting it, has a smaller pit than the latter cell and smaller stoppers. If the stoppers were not present, what would be the necessary upshot? The cell- contents would wander from one part of the plant to another, the main current flowing through the wider pit; consequently the main flow of proteid would be towards the tip of the main branch, and the plant would tend to form a single, greatly elongated thread. Of course, interposition of stoppers at the pit's mouth of the basal cells of lower-order branches must check the flow along such branches; but the net result will be equalization of the stream, and general distribution to all growing points of the necessary proteids will thereby be ensured. Onemay adduce several facts which seem to point to the existence of acallolytic ferment in sieve-tubes. Thus the disappearance of callus from the sieve-tubes of the rhizome of Phragmites com- munis described by Janezewski *—disappearance caused by simply placing the rhizome in a warm chamber during early spring—strik- ingly recalls the action of a ferment which is inoperative at low temperatures. Again, Russow T notes how in Pinus sylvestris and other woody plants the callus dissolves, usually about two years from the time of its formation, by a sort of corrosion, but in certain of the tubes it may remain more or less unaltered, some- times for ten years ; and there can be no doubt that side by side with open sieves, others may occur completely blocked with callus?. This one can readily understand by supposing that, for some reason or other connected with the metabolism of the tubes or of the companion cells, ferment has failed to make its appearance in the cells with callus-closed sieves. Again, we know from Russow's$ researches, that the callus of sieve-tubes of fallen leaves 18 not dissolved, and that small portions of starch also remain behind | ; the complete absence of eallolytie ferment in the one * * Comptes Rendus,’ 1878. t Sitzb. Dorpat. Naturf. Gesellsch. 1882. t Wilhelm, Siebróhrenapparatt. $ L.c | Briosi, Bot. Zeitung, 1873. seamen aaa a eT oa Bids IN YEGETABLE BIOLOGY. 519 case and ineffectiveness of amylolytic might explain this, or better, perhaps, the low temperature of the time of leaf-fall. And if there be a callolytic ferment in the tubes, it is easy to understand how, when the cool weather of autumn sets in, it would cease to be effective, and callus would then make its appearance upon the Sieves $. Considerations such as these led me to search for a proteolytic ferment in the sieve-tubes of the Lime-tree. I tried the method of Baranetzky as well as that of Von Gorup-Besanez, but in neither case with success. This is scarcely to be wondered at when it is understood how very small must be the quantity of callus, and hence of ferment necessary for its dissolution, even in an entire tree. Perhaps I did not choose a very favourable subject; and to Hop or Phragmites communis might answer the purpose etter. A Snort Nore on Macrocystis pyrifera, £g. Discovered in 1881 by Jeffrey Parker t and studied by Willt, Wille §, and F. W. Oliver ||, the sieve-tubes and trumpet-hyphæ of Macrocystis and Nereocystis, and the latter elements of other Laminarie, have been found with their contents in undoubted communication, Oliver has made the interesting discovery of a substance giving the ordinary reactions of callus in the sieve- tubes and trumpet-hyphæ of Macrocystis and Nereocystis €, and he has adduced evidence which leaves but slight doubt, if any at all, that the callus at least of the trumpei-hyphe is formed by swelling-up of the cell-wall. I have made many experiments With Macrocystis pyrifera, and in no case has the slightest trace of proteid reaction been observed; indeed the xanthoproteie is out of the question, for the callus rapidly dissolves in warin nitric * Sachs (Bot. Zeitung, 1862) finds the contents of the sieve-tubes to be alkaline; hence one would suppose that normal callus-digestion, if it really occurs, is of pancreatic, not of gastric nature, unless, indeed, temporary acidity make its appearance when the sieves reopen. t Trans. N. Zealand Inst. 1881, p. 562. t Bot. Zeitung, 1884. $ Kongl. Svenska Vet.-Akad. Handling. xxi, and Ber. d. deutsch. bot. Gesell, 1885. | Ann. of Bot. vol. i. . T Hick (Journ. Bot. 1886) has described continuity between the protoplasts of some Fucacee. It would be interesting to know if callus is developed in theso Cages, 520 MR. 8. LE M. MOORE’S STUDIES acid. In the sieve-tubes, on the other hand, there is upon the sieve-plates a substance which gives proteid reactions ; two of these are shown in Plate XIV. figs. ll and 12. It is evident that the callus of the trumpet-hyphe is altogether different in constitution from that of the Vegetable-Marrow, and probably it is produced by mucilaginous degeneration of the cell-wall ; moreover, it is unacted on by a peptonizing fluid. I cannot conclude without recording my obligations to the authorities of Kew and the Natural-History Museum, through my friends Mr. J. G. Baker and Mr. George Murray respectively, for their kindness in providing me with material of Ballia and Macrocystis. SUMMARY. 1. Vegetable-Marrow callus gives with great clearness all the three chief proteid reactions ; it is also dissolved by a peptonizing fluid; it is hence a typical proteid. 2.4 2. The callus cannot be pressed from the sieve, for (a) it 18 à proteid, (b) when it undergoes gastric digestion the sieve-plate 18 cleared, and is then left in its pristine condition. 8. The statements of the foregoing paragraph are supported by the entirely different reactions shown by wall and callus re- spectively. 4. Ballia-stoppers give all three proteid reactions, but are not attacked by a peptonizing fluid. They stain in the same way as does callus, except that they take a rich brown with iodine alone and are untouched by aniline-blue. d 5. The stoppers react altogether differently from the wall, an to a very large extent similarly to the cell-protoplasm. The evidence of differentiation within their substance, as seen when boiled in Millon's fluid, and the fact of their disappearance under action of Schulze's macerating fluid into plasma-balls which stil stain brown with iodine, speak for their similarity with cell- protoplasm. . 6. The group of bodies allied to proteids, such as mucin, nuclei, chondrin, &c., contains none which gives in a typical fashion & three proteid reactions. Hence the substance of Ballia-stopper® should not be classified with them. h 7. The callus of the Vegetable-Marrow has many of t d characters of, and should probably be classed with, the Coagulafe IN VEGETABLE BIOLOGY. 521 Proteids : the substance of the Ballia-stoppers most closely re- sembles Lardacein. 8. The function of the callus, both of the Vegetable-Marrow and of Ballia, is to moderate the flow of proteids and direct it so that all the growing-points shall receive their due amount of the necessary pabulum. That the stoppers function as sealers-up of the pit is shown by the coincident growth of pit and stoppers. 9. Many of the phenomena presented by the dissolution and renewal of callus-masses upon sieve-plates recall to mind the action of ferments. In these cases it is, in all probability, a eallolytie € proteolytic) ferment to which the effects are to be ascribed. This ferment has not yet been isolated. Supplementary Note. In the discussion which followed the reading of the above memoir, one of the speakers referred to Wiesner's * theory of cell-wall structure and to the observations relating thereto of Krasser + and Klebs f. It has been mentioned in the memoir that attention was paid to the proteid reactions of the wall, with the result that neither with the Vegetable-Marrow nor with Ballia did Millon's reagent or Copper Sulphate and Caustic Potash produce the slightest change of colour. Under these cir- cumstances, I did not see the bearing upon the question of the above-mentioned researches, and I fail to see it now in the case of Vegetable-Marrow callus, which, since it is soluble in a gastric fluid, can obviously have no genetic relation whatever with the wall. With Ballia, however, it is different, and I am obliged to Prof. F. W. Oliver for his reference to the subject. It is open to any one to argue thus :—“ True, the evidence brought forward that the stoppers are not formed by swelling-up of the wall is decisive ~ but when it is remembered how frequently the cell-wall contains either proteids or some nitrogenous substance or substances derived directly from destructive metabolism of proteids, it has not been proved that the stoppers, although they are deposited at the pit’s mouth by the protoplasm, may not consist of a carbo- * Bitzb. d. Wiener Akad. Band 93. t Béi Band 94. 1 Bot. Zeitung, 1887. - JOURN.— BOTANY, VOL. XXVII. 20 522 MR. S. LE M. MOORE'S STUDIES hydrate framework largely impregnated with substances giving proteid reactions, and this would account for the insolubility ina peptonizing fluid.” One method to resort to in such a case is to examine the stoppers for proteid after their sojourn in the gastric fluid; if the proteid reactions now fail, obviously the proteid will be merely an infiltration-substance. But if these reactions be now given, this may be due either to insolubility of the proteid in the fluid, or to the presence of nitrogenous decom- position-products. On referring to my notes, I cannot find that I actually tested for proteid after the action of the peptonizing fluid; my belief is that I did—indeed, it is obviously natural so to do: it is possible, however, that I may have relied upon the evident intactness of the stoppers, their form, size, and refrin- gence being precisely the same after as before the action. There is, however, a crucial test applicable to the case ; but before going into the matter further, it would be well to mention in a few words what are the views of Wiesner and his followers, and what are the objections which have been urged against them. Wiesner's dermatosome theory of the cell-wall depends upon two facts :—(1) capacity of the wall to break up, on suitable treat- ment, into small definite particles or dermatosomes ; (2) the presence of living protoplasm (dermatoplasm) in the cell-wall. Thelatter Wiesner holds to consist at first of protoplasm alone, and he contends that as long as the wall is growing it contains dermatoplasm. The cell-wall, which has a reticulate structure, consists of altered microsomes of the protoplasm, or dermatosomes ; these are united by delicate strings of dermatoplasm, out of which are continually being formed, as long as the cell-wall is growing, at first new microsomes and then dermatosomes. Thus the cell- wall is, as long as it grows, a living member of the cell. Wiesner coneludes that proteids exist within the walls of young cells, from the behaviour of the meristem at the vegetative point of the stem and of the cambium and phellogen of various plants. After acting upon these with a peptonizing fluid he asserts that he can get the cellulose reaction, not before. Upon this subject we have the observations of Mulder *, who, half a century ago, suspected the presence of proteid in the cell-wall, although Wiesner has himself shown that the facts upon which Mulder relied are other- * Berzelius, Jahreeber. 1840, p. 649; quoted by Krasser, Sitzb. Wien. Akad. Bd. 94. "e 1» F M PET IN VEGETABLE BIOLOGY. 598 Wise explieable. More in point is the fact that Richter* showed - the probability of the existence of proteid in the wall of fungus- cells, and that N enckit has found proteids in the wall in the Bac- teria (mycoprotein and anthraxprotein) ; while Forssell f, shortly after Wiesner, by means of Raspail’s and Millon's reagents, dis- covered proteid in the wall of certain Lichen-hyphe (Lobaria pul- monaria, Peltigera canina) and Algæ (Gelidium cartilagineum, Ecklonia baccata, &c.). We owe to Krasser an extended research upon this subject. He first made a careful study of all the non- proteid bodies which glve one or more of the proteid reactions : thus he finds .— . ; The Xanthoproteic reaction, given by Tyrosin, certain resins and alkaloids. bi AMillon's reaction, given by starch, cotton-wool, and gum-ara T (as Millon himself found), Tyrosin, Phloroglucin, € i Thymol, Vanillin, N aphthol, Nitrobenzole, and id: 10 oxyacids. . Deen reaction, given by oil, certain fats, Tyrosin, Phenol, Thymol, Naphthol $. . . . éd test (blue colour with Molybdic Acid and pons cilphate) given by Glyoarin, yaeta: degen cane-sugar, Phenol, Phloroglucin, Thymol, Vani, Coniferin. . Krasser’s Alloxan reaction (purple-red colour with watery solution), given by Tyrosin, Aspartic Acid, nung, The upshot of Krasser's labour was the conviction t at Fins protoplasm is present in the cell-wall of Fungi (hyphe Lichen id porus sulphureus and P. fomentarius and of sever " ) Algæ (Chondrus crispus, Oladophora and Edogomum Ke (failure with Chetophora, Spirogyra, Zygnema), C. Sa of Pha- of Polytrichum, several Vascular Cryptogams, meris ) epi- and Derogams (vegetative point, pericambium, phellogen), ep * Sitzungsb. d. Wiener Akad. Band 83. h. chem. Gesell. T Journ. f. prakt. Chemie, Neue Folge, xx., and Ber. d. deutse . Em jen, Engl ahrg. xvii. ; both quoted by De Bary, ' Vorlesungen über Bacteri transl. p. 4, Sc, Nat. 6 sér. 1 Sitzungab, d. Wiener Akad. Band 93. See also Frémy, Ann. 13. : itute Molisch's -© $ Precipitates are formed in these last two ai, vii Molisch, Sugar-reactions (violet with naphthol, rusty-red wi : Anz, d. k. Akad. d. Wiss. Wien, 1886, p. 97). 202 * 524 MR. S. LE M. MOORE'S STUDIES hypodermal tissue, collenchyma, tendrils (Bryonia, Cobea, Cucu- mis, Vitis), endosperm cells (occasionaliy). Moreover, by taking proper precautions, he claims to have found protoplasm also in the xylem- and phloem-fibres of several plants. These views of Krasser and Wiesner have met with severe handling on the part of Klebs* and Fischer+. Klebs holds that, even if there be proteid in the cell-wall, which has not been proved, there is no justification for the supposition that we have to do with organized living protoplasm. He favours Strasburger and Schmitz'si theory of the origin of the cell- wall viz. that the peripheral layer of protoplasm is directly changed into cellulose; and as an example of the rapid way in which this transformation takes place, he cites the membrane around ejected plasma-masses of Vaucheria. This cannot be living protoplasm, for, by contracting the latter in a weak solution of syrup and staining with Congo-red—a dye which is not taken up by protoplasm-—we learn that cellulose has been formed. He also, by employing Lów and Bokorny's $ alkaline silver-solution method, which Krasser himself had par- tially relied upon, finds that, if Krasser’s application of the method be correct, there is unmistakable evidence of protoplasm in the wall of the pitted tracheides of young Maize plants, while the other cells react very much less clearly! Fischer believes it impossible to distinguish proteid by micro-chemical means alone; he says that morphological and developmental facts must always be adduced, and he submits Krasser's results with Bromeliaceous leaves to very damaging criticism. He shows, too, that it is not the presence of proteid which hinders the cellulose reaction in the case of young membranes, but that time is required to stain them blue with Schulze’s solution. Also that in some cases (sections of Nidularium leaf, of the cotyledon of the bean, and of endosperm of Ricinus) the cell-walls take the same colour with Millon’s reagent after, as before, the action of a peptonizing fluid. He thinks it probable that not proteid, but some substance resulting from direct destructive metabolism of proteids, probably Tyrosin, gives the reactions upon which Krasser and Wiesner rely. To * Bot, Zeitung, 1887. + Ber. d. deutsch. bot. Gesellsch. 1887. 1 Strasburger, Ueb. d. Bau und d. Wachs. de Zellhaüte ; Schmitz, Sitzungsb. . d. niederrh. Gesellsch. in Bonn, 1880. $ Die chem. Kraftquelle im lebend. Protoplasma : München, 1882. IN VEGETABLE BIOLOGY. 525 these strictures Wiesner * has replied, but without materially advancing the theory. Is some such substance present in Bailia-stoppers? Bearing in mind the identity in proteid-reactions shown by stoppers and protoplasm (omitting, of course, the brighter colour with Millon’s fluid taken by the stoppers), one is inclined to say no. In order to decide the question, however, experiments have been made with fluids in which the non-proteid substances giving Millon’s reaction are soluble. l It may be premised that Phloroglucin cannot be present, for Hydrochloric Acid gives no red colour to the stoppers. Of the bodies that remain, Nitrobenzole, Phenol, Thymol, and Naphthol are readily soluble in alcohol and ether. Accordingly pieces of Ballia were allowed to lie in strong alcohol for 36 hours, and, on Dow testing with Millon’s fluid, the bright red colour was at once got. As for Tyrosin, it is soluble in acid and alkaline solutions ; but on testin g with Millon's fluid after 36 hours’ action of a strong Solution of Hydrochloric Acid, the red colour was easily yielded. Vanillin is soluble in water; 90-100 parts are required at 14°, and 20 parts at 75°-80° +; but on soaking Ballia-plants in water for three days and afterwards boiling for five minutes, Millon’s Teaction was as readily obtained as ever. There remains the small and obscure group of aromatic oxyacids; but I cannot see the slightest chance of the proteid reactions of the stoppers being due to them, [Supplementary Note to Memoir on Apiocystis Brauniana, Nag. in Journ. Linn. Soc. Bot. xxv. (1890), pp- 362-380.—In the Memoir on Apiocystis a few notices of the distribution M this alga escaped me. On referring to the recently issue de elaborate t Sylloge Algarum ’ of Dr. De Toni (Padin one Appears that Apiocystis has been found in Sweden, Russia, i r , and in the United States. There is also a paper in the ‘Ann d in Sciences Naturelles’ (sér. 7, tome vii. p. 158) by M. Dangeard, in Krasser, Bot. Zeit. 1888, * Ber. d. deutsch. bot. Gesellsch. 1888. See alao im’ Ge e vise Bot. Kohl (Bot. Centralbl. vol. xxxvii.) and Krabbe (Pringsbeim’s Jahrb. f. xviii. Pp. 352-354) write against Wiesner’s theory. ; ses - t Roscoe and Schorlemmer, ‘Treatise on Chemistry,’ vol. iii. part 4, p. 354. - 526 MR. a LE M. MOORE’S STUDIES which mention is made of an examination of Apiocystis, so that the alga would appear to occur also in France. Professor Bower tells me that he found Apiocystis while working, some years ago, in the late Professor de Bary's Laboratory at Strassburg. Other British localities are Aberdeen (Roy) and Yorkshire (West), ac- cording to Mr. A. W. Bennett in Journ. Roy. Micros. Soc., Feb. 1890. See also Mr. West in Journ. R. Micros. Soc., June 1890. Professor Hansgirg’s memoir “ Ueber den Polymorphismus der Algen " (Bot. Centralbl. Band xxii. p. 200) was overlooked by me. Hansgirg imagines a number of algal types, including such ap- parently distinct ones as Nephrocytium, Oocystis, Dictyospheria, Characium, Hydrianum, &c., to be mere pleomorphic forms of various Chetophoracee, Siphonocladiacee, Ulvacee, Ze, He is unfortunate in including Apiocystis in this list; and one can scarcely suppose that even careful research will confirm all this botanist’s ideas.— 8. A I EXPLANATION OF PLATE XIV. Fig. 1. Vegetable-Marrow. Sieve-tubes and companion cells: xanthoproteic reaction (reaction of wall not shown). 450. 2. Ditto. Millon's reagent. 450. 3. Ditto. Oopper Sulphate and Caustic Potash (the dirty-pink colour). x 450. - Ballia callitricha, Ag. Millon’s reagent. 450. . Ditto. Xanthoproteic reaction. 450. - Ditto. Copper Sulphate and Caustic Potash. . - Ditto. Stoppers after 30 hours’ action of the peptonizing fluid. Pit and connecting-protoplasm well seen. (Drawn with mag. of 600 diam. and enlarged.) 8. Ditto. Shows effect of boiling for one minute with Millon'sfluid. The colour has flown and granular precipitation is visible in the stoppers ; protoplasmic reticulum still visible. 600. . Ditto. Shows action of Schulze’s macerating fluid: (a) the stoppers have melted into the two balls of plasma which still remain at the pit’s mouths ; (5) plasma separated from pit, showing the latter quite empty. X450. Vegetable-Marrow, Effect of a peptonizing fluid: the sieves are quite clear, and no trace of callus is visible. 450. Macrocystis pyrifera, Ag. Callus upon sieve-plate of a sieve-tube: xan- thoproteic reaction. x 600. - Ditto. Effect of Millon's reagent upon callus of sieve-tube. 450. To En Hm 10. 11. a 8 | 5 E: Ln. Soc. Journ. Bor. Mat, XXVII P]. 14. > IN VEGETABLE BIOLOGY. 527 Stupizs rN VraxrABLE Broroey.—VII. Some Microchemical Reactions of Tannin, with Remarks upon the Function of that Body and its Excretion from the General Surface of Plants. By Spencer Le M. Moons, F.LS. [Read 17th April, 1890.] Axour eighteen months ago, happening to have by me some of the fluid proposed by Nessler many years since as a test for ammonia (solution of potassium iodide saturated with mercuric iodide to which caustic potash is added), the idea occurred to try whether this would be of any value in the microscopical study of tannin by the botanist. It was found that with solutions of tannin a dark-brown precipitate is thrown down by Nessler’s fluid; hence one would expect to find the brown colour imparted to tannin when in the cell of a tissue. Since that time many experiments have been made, with the result—already briefly announced in ‘ Nature ’*—that the fluid in question is likely to prove a valuable auxiliary to the already long list of tannin- Teagents at the botanist’s disposal. Just as with iron salts the effect produced is not always the same, so too with Nessler's fluid. That it may not be thought that some other substance or sub- stances besides tannin gave the reactions hereafter to be de- scribed, it must be remarked that great care was taken to accu- mulate evidence, both positive and negative, as to the nap being due to tannin in one of its forms, and to tannin alone. , is, however, unnecessary to dwell upon Mi matter, since it has already been set forth in the letter to ‘ Nature. Besides the brown colour spoken of in the letter, farne research has brought out a good deal of difference in the behaviour of Nessler’s fluid towards tannin. Three kinds are to be distinguished, viz. :— , , I. Tannin giving an immediate brown precipitate, occasionally with decided brown-pink tendency. . : IL Tannin giving a yellow colour, quickly becoming red-brown, and, finally, a cold brown precipitate. III. Tounin giving a yellow colour, the yellow lore readily diffusing through the cell-walls into the sarron » $ fluid, thus leaving the cells colourless after a varying lap time. - * Vol. xli. p. 585. 528 MR. 8. LE M. MOORE’S STUDIES Group I. (Immediate brown precipitate.) In the fundamental tissue of Musa sapientum run numerous large tannin-saes in the neighbourhood of and parallel to the course of the vascular bundles $. Nessler’s fluid immediately turns this tannin a fine rich brown. Epiderm (upper and lower) of leaf of common Primrose. Numerous tannin idioblasts are scattered among the ordinary cells, which latter are well provided with chlorophyll. Rosa canina. The tissues in the neighbourhood of the apical meristem of the stem are richly provided with tannin imme- diately giving the brown colour. The same is the case with the garden Rose. Epiderm of Grevillea robusta. Tannin occurs plentifully in the cells overlying the vascular bundles on the lower side of the leaf, also in a few cells in their immediate neighbourbood : else- where tannin is restricted to the two subsidiary cells which lie on each side of the stomatal guard-cells. Group II. (Yellow colour, changing through red-brown to brown.) A good example of the occurrence of this kind of tannin is the fundamental tissue of the young stem of Aucuba japonica. At first the colour is a pronounced yellow; but in a few minutes this gives place to brick-red, which is itself supplanted by a warm brown. After a few days in glycerine, the colour of the tannin- precipitate in such a preparation changes to a cold brown. Group III. (Readily diffusible yellow colour.) This has been frequently met with. As instances, may be cited the following :— Leaf-scales of scape of Petasites vulgaris and Tussilago Farfara. lvy. Fundamental tissue of young stem and the large stellate hairs clothing it. Dandelion. Fundamental tissue of underground stem. Potentilla Anserina. Hairs thickly covering the plant. Hairs on stipules of Rosa canina. * De Bary, Vergl. Anat. p. 160 (Engl. transl. p. 153). IN VEGETABLE BIOLOGY. 529 Chestnut. Abundance of this kind of tannin in the cells of the bud-scales. The chief value of Nessler's fluid to the microscopist is that, apparently in consequeuce of the rapid diffusibility of caustic potash, its action in showing up any tannin-containing cells in a tissue is very quick. This valuable property is alluded to among the further details to be now given. Group I. Primrose epiderm.—To see the tannin in this plant, the easiest way is to tear off pieces of epiderm from the under- side of the leaf, and mount them in the testing-fluid. The strips should include some cells overlying a vascular bundle. With Nessler’s fluid one immediately sees that tannin idioblasts are present in some numbers in the elongated cells overlying the vascular bundles, and from the very dark, frequently quite black, colour of the precipitate, the richness of these elements in tannin is manifest. Numerous tannin idioblasts are, as has already been Mentioned, scattered among the ordinary epidermal cells: the latter contain a nucleus and numerous chlorophyll corpuscles, which, contrary to what is usually the case, are about the same size as those of the guard-cells. The guard-cells give no sign of tannin, and it is rather rare to find tannin idioblasts bordering on them $. The hairs covering the leaves consist of a simple row of cells, and show a curious disposition of the tannin ; as a rule, the alternate cells are tannigerous, the intervening ones have chlo- Tophyll and are without tannin; the proximal cell of the hair may have tannin, or be a chlorophyll-cell. Besides Nessler’s fluid, the only ordinary reagent able to show up the tannin of the Primrose without any loss of time 18 Osmic acid in 1-per-cent. solution, which causes the well-known inky-blue precipitate to be thrown down in the tannin elemen s. As an illustration of the long time required to stain the tannin- cells in such a case as Primrose epiderm, if the ordinary tests are used, the following Table is subjoined. * At an earlier stage the tannin-idioblasts contain chlorophyll; n they - at once picked out from their neighbours on account o trois ro of their chloroplasts; afterwards the chlorophyll disappears 530 MR. 8. LE M. MOORE'8 STUDIES Time during Reagent. which experiment Effect produced. lasted. — Iron acetate ...| 12.55 p.m. to 4 P.M. Tannin not yet completely shown up. Ferrous sul} 2.45 P.M. to 4 P.M. | Effect imperfect. Potassium \ bichromate. | 11.30 a.m. to 4 P.M. but many not yet stained or but slightly. Many cells beautifully coloured ; many not coloured at all. Addition of ammonium chlo- ride to the molybdate caused | ` no perceptible difference. | A pale yellow-pink or pale Ammonium \ molybdate. l Many idioblasts well stained, ( 11.15 a.m. to 4 p.m. |4 | | Copper acetate. | 1.20 p.m. to 4 p.m. brown colour in most of the tannin-cells. 3 pM. to 4 P.M. Grey precipitate in a few ofthe —— ° tannin-cells. On examining, next morning, strips of epiderm placed over- night in the above solutions, the effect characteristic of each was well seen. In the subsidiary cells of Grevillea robusta the tannin shows ` up immediately with Nessler’s fluid, as also with osmic acid. Some of the ordinary reagents, such as iron acetate, will act directly upon a few of the cells ; but many cells remain uncoloured, and several hours must elapse before they stain. It is further to be noted that in every case in which an immediate brown pre- cipitate is given with Nessler’s fluid, a blue-black precipitate is thrown down with iron salts. Group II.—We have seen that tannin in Rosa canina gives à brown precipitate with Nessler’s fluid; but the hairs upon the young leaves take at first a yellow colour, and in some of them this passes into a brown. It would hence appear that we have two kinds of tannin in these hairs. Nor is this matter for wonder, since Dufour * finds both iron-blueing and iron-greening Ss Soc. Vaud. 1886. Abstracted in Journ. R. Micros. Soc. 1887, P ^s D IN VEGETABLE BIOLOGY. 531 tannin side by side in the same tissue, viz. the epiderm of Sedum Telephium. Group ITI.—The use of Nessler's fluid is especially valuable here, inasmuch as other reagents may not act within a reason- able time, Strictly speaking, however, it is the caustic potash which produces the effect ; indeed any alkali will give the yellow colour with this form of tannin, as bas been known for many years. So far as T am aware, Wiesner * was the first to draw attention to this matter, and he also noticed that this alkali- yellowing tannin gives a green precipitate with iron salts. I can confirm Wiesner on this point, as in every case in which the yellow colour has been given with Nessler's fluid, iron salts have thrown down a green or greenish-yellow precipitate. The chief point to note is the light yellow colour immediately given with Nessler's fluid , 4s also with alkalies, and (what has not, it is believed, been observed before) the diffusion of this coloured matter through the cell-walls, so that after afew minutes the piece of tissue under examination is surrounded by a yellow halo easily visible to the naked eye. Take, asan example of the occurrence of this tannin, the scales of the scape of Petasites vulgaris; the following re- marks apply to the epiderm alone, but tannin occurs elsewhere 48 well Nessler's fluid immediately brings out the yellow colour in all the cells except the guard-cells of the stomata, and the tissue iS soon rendered colourless for the reason just given. Osmic Acid within three minutes stains almost all the cells a pale slaty Srey, a colour which osmic acid always gives, at least as far as one’s observation goes, with the iron-greening tannin. Ferric chloride immediately gives a dark green precipitate in all the cells, and potassium bichromate acts pretty fairly after some time, half an hour at least. Many cells are at once stained the beautiful and characteristic colour with ammonium molybdate ; but even after two hours only groups of cells are coloured. Again, the large stellate hairs closely set upon the axis and appendages at , the growing-points of the Ivy—the ordinary reagents are not nearly so satisfactory in this case as in the last, for, whereas the action with Nessler’s fluid is immediate, with ferric chloride a couple of hours must elapse before the green precipitate is thrown down. Ammonium molybdate will colour a few of the hairs within ten minutes, but many remain unaffected a much longer * Bot. Zeit. 1862. a ig 532 MR. 8. LE M. MOORE’S STUDIES time; while for the slaty colour with osmic acid one must wait half an hour or so. Potassium bichromate also is slow in its action. The close relation between the colouring-matters which occur dissolved in the cell-sap, and are known under the general term anthocyan, is admitted by most physiologists *. This is what happens when cells containing anthocyan (e.g. epiderm of red- cheeked apples, hairs of Escallonia macrantha and of Petasites vulgaris scales) are acted on by Nessler's fluid. The red colour is almost instantly turned a fine deep green, which soon passes into yellow and this into brown. If caustic potash alone is used, the effect is the same except that the brown stage is not reached. Here, too, one sees how valuable an aid Nessler's fluid is likely to prove: for instance, De Vries t, in his study of the contents of Drosera tentacle-cells, used Méll’s test for the discovery of tannin, and this involved waiting for twenty-four hours, whereas with the new reagent his inquiry would have been answered in a few minutes. Failure might, perhaps, be deemed probable with cells con- taining much free acid, since the action of Nessler’s fluid is due in large measure to an alkali. But, so far as concerns two of the commonest acids occurring in plants, this is not the case, for on charging a solution of tannic acid with either citric or oxalic acid, the precipitate is still obtained with Nessler’s fluid. Added to a solution of tannin or tannic acid Nessler’s fluid throws down a precipitate which is momentarily of a reddish colour, then changing to brown. Other reactions are as follows :— With Gallie Acid it gives a grey-green precipitate. » Pyrogallol » a brown precipitate immediately (no momentary red). a deep green precipitate, soon turning to greenish brown. » Protocatechic Acid ,, a dirty-green precipitate. » Pyrocatechin » These precipitates are probably all of them oxidation-products; * On this point see, for instance, Wigand, Bot. Zeit. 1862, and Forsch. aus 169 d. bot. Garten zu Marburg, Heft ii. (1887) ; Wiesner, Bot. Zeit. 1863; Went. \/ Jahrb. f. wiss. Bot. Band 19; Pick, Bot. Centralbl. 1883; Dennert, Bot. Cen- tralbl. 1889, t Bot. Zeit, 1886. IN YEGETABLE BIOLOGY. 533 but further information on this subject can scarcely be expected from a botanist., A few words now, before proceeding to speak of some other tannin-reactions, about the “ glucoside or glucoside-like sub- stance" recently found by Haberlandt * in the drops of fluid exuding from cut surfaces of Mimosa pudica, and called by former authors “ water-drops." This substance is met with in certain large elements scattered through the leptom, which Haberlandt believes is the path taken in the conduction of a stimulus from one part of the plant to another; it is strongly acid, and is held by Haberlandt, and apparently with reason, to be the substance by whose osmotic activity is brought about the relatively high degree of turgescence the abolition of which on sectioning results in exudation of the drop. With iron salts a purple colour is taken by it, as Haberlandt shows, and, on heating with acids, a body- reducing copper oxide is formed. Had Haberlandt applied some other tests he would have found that not only is the body in question a glucoside, but that it is tannin or tannic acid. Thus ammonium molybdate gives a decided yellow, and potassium bichromate a brown colour to the crystals which rapidly form in the exuded drop, and the development of which is a very pretty sight. Nessler's fluid is less satisfactory : a well-marked yellow solution is obtained with it, and a slight brown residue which has to be carefully looked for. It is neces- sary, in applying these reagents, to use them in very small quan- tities, otherwise the result sought for will not be reached ; 80, too, with iron salts, it is only at the junction of tannin with salt that the purpling occurs. It may be stated as a general rule, that addition of caustic potash to a tannin-reagent quickens its action. Thus if it rein- forces ammonium molybdate an immediate, though faint yellow Colour is given to the idioblasts of the Primrose and to the sub- Sidiary cells of Grevillea robusta. But inasmuch as caustic Potash alone will produce a closely similar effect, it is not aon whether the molybdate cooperates in this. With caustic a and iron sulphate an almost immediate purple or purple- ww precipitate is thrown down in the cells of the Primrose ; but a experiments with these two bodies have proved unsatisfactory on Account of the precipitate caused by their coming in contact. * ‘Das Reizleitende Gewebesystem der Sinnpflanze, Leipzig, 1890. fe 7 ^ ^f. 534 MR. 8. LE M. MOORE'S STUDIES More useful is a mixture of iron acetate and caustic potash, which rapidly colours tannin a brown-pink. With copper sul- phate and potash also the effect is marred by the presence of a precipitate: the tannin idioblasts of the Primrose are coloured straw-yellow by this means, and with copper acetate the action, formation of a greyish precipitate, is uncertain. Nessler’s fluid without the mercuric iodide, d. e. caustic potash with potassium iodide, gives, in the few cases in which it has been tried, a quickly- got pale brick-red colour. This is a reagent which ought to be further studied, inasmuch as it is free from an objection to the employment of Nessler's fluid founded upon the poisonous nature of mercuric iodide. Addition of caustic potash to potassium bi- chromate is of no use. A modification of Möll’s test, in that iron sulphate was substituted for acetate, turned out satisfactory, at least in the case of the Primrose epiderm, since after overnight action of copper acetate the sulphate threw down a precipitate of iron tannate directly, whereas the acetate did not do so for some ten minutes. What one is inclined to consider a fact of some importance to plants, has come to light in consequence of these experiments with tannin. It has been already mentioned that one form of tannin turns yellow with Nessler's test, and that this yellow fluid diffuses out of the cell. It must be confessed that the meaning of this quite escaped me until I came upon Wiesner's * paper, in which it is stated that alkalies generally will colour the iron- greening tannin a bright yellow. The importance of the diffusi- bility of this alkali-yellowing substance is seen when the constant presence of ammonia in the air and in rain-water is remembered. But it is necessary to inquire whether the minute quantities of ammonia found in the air and in rain-water are sufficient to convert the tannin into a yellow diffusible substance. To answer this question a few experiments have been made with the Ivy; in the course of these it was ascertained that exceedingly weak solutions are capable of acting in the above capacity. Thus a 1-per-cent. solution of ordinary * liquor ammoniz ” of the shops— itself a diluted solution—gives a pronounced yellow colour to the hairs, epiderm, and fundamental tissue of the young Ivy stem, * Bot. Zeit. 1862. Pfeffer (Ber. d. deutsch. bot. Gesell. 1886) has noticed that the tannate formed by action of methyl-blue upon a tanrigerous tissue diffuses out into the surrounding medium . IN VEGETABLE BIOLOGY. 535 and even with a 0'5-per-cent solution, to which Nessler's fluid Imparts a scarcely visible yellowing, the yellow colour is clearly enough seen in many of the hairs and epidermal cells. Let us now see how this operates in nature. A wet day comes, and the leaves and stems of herbs, shrubs, and trees are flooded with water containing ammonia in solution; this must obviously cause diffusion of epidermal aud trichomal tannin in the form of the yellow substance in question, and one can scarcely resist the conclusion that the whole surface of the plant—at least in the case of herbaceous plants—must play under these circumstances an excretory rôle. And since dew contains a much higher pro- Portion of ammonia than does rain-water, the argument applies With yet greater force to the nightly vicissitudes to which vegetation is exposed. But the case becomes even stronger when we consider under- ground parts. Among other ways in which ammonia is accumu- lated in the soil, we know that decomposition of nitrates and of nitrogenous organic bodies is ever going forward in consequence of the activity of ammoniopoietic micro-organisms, and it can scarcely be questioned that this soil-ammonia must be an efficient agent in the excretion of tannin. Numerous as are the writers who have treated of that obscure subject the function of tannin, one can scarcely admit that any of them has made us much the wiser for his pains, though this is scarcely a matter for special remark, in view of the abstruseness of the theme *. Some would have us believe that the tannins, * Among the chief authors of recent memoirs on tannin are the following ; in each case the writer’s views as to the function of tannin are succinctly stated within brackets :—Sachs C Vorlesungen, no. xi.) [Excretory and estive) Kutscher C Flora, 1883) [Exeretory and assimilative]. Gardiner (Proc. Cam , Phil. Soc, vol. iv. p. 888) [Excretory]. Vines (* Physiology of Plants,’ p. "A [Decomposition-produot, but indirectly assimilative]. Kraus (Ber. Sitzb, naturf. Gesell. Halle, 1884; abstr., Journ, R. Micros. Soc. 1886) [Reserve Substance]. Westermaier (Sitzb. d. Berl. Akad. 1885 and 1887) [Assimnilative} De Wildeman (Bull. Soc. Roy. Bot. d. Belgique, 1886) —— gt uni ] (Bull Soc, Vaud. 1880; abstr., Journ. R. Micros. Soc. 1887) [ Ty}. Schulz (‘ Flora,’ 1888) [Assimilative]. A. Fischer (Bot. Zeit. s De d. Hillhouse (* Midland Naturalist, vols. 10 & 11) [Exeretory]. uid "iy deutsch. bot. Gesell. 1888) [Assimilative]. Stahl (P animals]. G. Kraus in Jen. Zeitschr, f. Naturwiss. 1888) [Protective apu 1889; iy Journ ( Grundlinien zu einer Physiologie d. Gerbstoffs, Leipzig, est anim als] R. Micros. Soc. 1889) [Antiseptio, excretory, and protective agains OST NR 536 MR. S. LE M. MOORE'S STUDIES such ready yielders of that important translocation-product glucose, are entirely excrementitious ; others hold them to be themselves translocation-products ; while a middle course is steered by yet others, who think that both functions, assimilative and excretory, are performed by these bodies. An antiseptic rôle has also been ascribed to tannin, and, on account of its bitterness, it has been supposed to protect plants against the depredations of animals. I cannot help thinking that it is rash to argue, because direct participation of tannin in metabolism is not discoverable, that therefore it is an excretory substance. An important part may possibly be taken by it in the lignification of cell-walls; but it will not be necessary to treat of this matter except to refer to the apparent presence of glucosides in lignified cell-walls*. In tannin we have a glucoside which is present often in great quantities in the plant. tissues : it undoubtedly moves from one part to an- other, and its production depends upon conditions favourable to metabolism. On the other hand, it cannot be distinctly shown to participate in metabolism. What, then, can have become of it? Is the glucoside in lignin formed afresh when there is already plenty of tannin in the plant? And even with regard to the iron-greening tannin, which we have seen reason to believe is largely excreted from the surface of plants f, and which is doubt- less primarily a waste product, it must be remembered, since Nagelit has found that tannin can function as the source of carbon to fungi, that the metabolism of fungi may perhaps involve katabolism of quasi-useless bodies, such as tannin, and pro- duction of substances capable of being used up in the nutrition of higher vegetation. Busgen (Jen. Zeitschr, f. Naturwiss. 1889) [Protective against animals: share, if any, in constructive metabolism not yet proven]. ^ See Vines, ‘ Physiology of Plants,’ Lecture ii., with the bibliography there given. T The storing up, by the temporary hairs of glabrate organs, of tannin formed in the meristem may also be a means whereby the plant is enabled to rid itself of the glucoside. The hairs of the Ivy, for instance, are large and can contain a relatively great quantity of tannin, and they soon drop off. But Schenk (‘ Ver- gleichende Anatomie der submersen Gewiichse,’ quoted by Stahl in Jen. Zeitschr. f, Naturwiss. 1888, p. 594) regards the caducous tannin-filled hairs of Cerato- phyllum as a special protection for the young organs. 1 Sitzb. der k. bayer. Akad. der Wiss. 1880, p. 339. See also Van Tieghem, * Traité de Botanique; p. 542. IN YEGETABLE BIOLOGY. 537 In the epiderm of the Primrose we find tannin idioblasts scattered among the chlorophyll-containing cells, and we have already seen how, in earlier states of the leaf’s development, these idioblasts are only distinguishable from the cells among which they lie by their containing fewer chlorophyll corpuscles, although even now they have much tannin. What is it which determines whether a cell shall become an idioblast or remain a chlorophyll cell? What happens is probably this:—One con- sequence of cell-metabolism is the formation of tannin, and this tannin is passed on from cell to cell either by diffusion through the wall or by active movements of the connected protoplasts *. The cells, then, besides manufacturing tannin, are continually receiving it from other cells and passing it on ; the tannin-streamst will thus obviously tend to converge towards certain points, and at these points tannin will accumulate. Accumulation of tannin is unfavourable to metabolism, and after a time the cell will de- generate to a mere passive recipient of tannin from other cells. But there must also be lateral movement in the epiderm to the idioblasts overlying the vascular bundles, for there is almost always more tannin in these idioblasts than in those situated in other parts of the epiderm 1, and it is this evident movement of tannin towards the conducting tissues which would seem to argue for it some other than a merely excretory rôle. Whether the tannin is of any service to the plant while in the idioblast is not at all clear. Of course it may preserve the leaf from the attacks g injurious animals (insects &c.); but when one recalls Haberlandt 8 recent discovery relative to tannin (or rather tannic acid) as being the osmotically active substance in Mimosa pudica, and connects with it the fact that the tannin-idioblasts of the Primrose are almost always after a time somewhat, and often markedly * Either method is possible; for on the one hand Vesque (‘ Comptes Rendus, 1886 ; abstr. Journ. R. Micros. Soc. 1887) finds that water passes with great ease from one epidermal cell to another ; and continuity of protoplasm is often to be seen with exceptional clearness in the epiderm (e. g. Schaarschmidt, T4 Nóv. Lapok, 1884, abstr. Journ. R. Micros. Soc. 1885; and Tangl, Sitzb. d. k. - Wiss. Wien, 1884). t That there are in streams is proved by the existence of the so-called tannin-bridges (vide Moeller, Ber. d. Deutsch. bot. Gesell. 1888). NN $ This point was enunciated in the writer's second memoir on ZS th Protoplasm ” (Journ. Linn. Soc., Bot. xxiv. p. 377). Some months ^ d e reading, but a little while before the publication of that memoir, Moeller (l. c.) made the same discovery. 4 LINN. JOURN.—BOTANY, VOL. XXVII. =P 538 STUDIES IN VEGETABLE BIOLOGY. larger than the chlorophyll-cells, one is inclined to ask whether the presence of these idioblasts may not be connected with the closing movement of the stomata. And, lastly, why may not tannic acid have some general relation to the turgescence of cells? SUMMARY. 1. Nessler's test for ammonia is a valuable aid to the botanist in detecting with certainty and rapidity the presence of tannin and tannic acid in plants. Other fluids, having eaustic potash for a basis, are also good reagents for tannin. 2. Two chief kinds of tannin are to be distinguished, according to their behaviour with Nessler’s fluid : (a) the iron-blueing tannin strikes brown with the fluid, (b) the iron-greening variety is turned yellow by it. 3. The yellow substance just mentioned readily diffuses through the cell-wall; this effect is to be ascribed to the caustic potash, for alkaline solutions, even the weakest, will act in the same way. Here we have a provision, by the aid of rain, dew, and activity of aoil-organisms, for the excretion of tannin from the general surface of all plants containing this form of it. 4. In addition to the functions hitherto ascribed to tannin, Haberlandt’s recent discovery with reference to the ‘ water-drop’ exuding on section from Mimosa pudica, renders it possible that tannic acid may have a more general relation to the turgescence of cells. Moreover, tannin is most likely used up in the lignifi- cation of the cell-wall. 5. The diffusible tannin, although primarily excretory, and the non-diffusible kind when occurring in shed organs, may yet, in view of the fact that tannin can act as a source of carbon to fungi, have some indirect connection, vid the nutrition of Sapro- phytes, with the metabolism of green plants. INDEX. Abies Veitchii 236-241, 249, 252 254, 272, 974. : Webbiana, 236, 237, 254. (Picea) excelsa monstrosa, ftnote 283. Abietinez, 236, 279-290, 300-326. Abrus precatorius, Linn., 10, 30. Acacia Farnesiana, Willd., 10, 18, 34. Acalypha Noronhe, Ridley, 59. Acanthophora multifida, Lamz., 76. Thierrii, Lazz., 76. Acanthophyllum rosulatum, Hook. & Abies, my 236, 274-328. amabilis, 247, 254, 263, 272 apollinis, 236, 237, 240, 257. dd samea, 236 2 SS 2» 236, 237, 240, ftnote 241, lida, 249, 254, 263, 972. 28 braehyphylla 272. ` ` racteata, 249 ő acte , ftnote 255, 272, Brunoniana, 320. cephalonica, 236, 237, 240, 249, a4. 272, 285. iicica, 249, 954, Arn., 493. concolor, 254, 272, Acanthospermum hispidum, DC., 42. Acer, 126, 134. Davidiana, 249. excelsa, 128, 129, 131, 229, 777, var. monstrosa, 229. —— Var. viminalis, 229. "a var. virgata, d excelsa Cranstoni, SI Zë firma, 249, 254, 263. Fortunei, 254. eraser 250, 254, 272. ndis, 236, : 257 29 240, 247, 949, 254, homolepis, 249, 254, 272, 273, campestris, 202. Achnantheæ of Fernando Noronha, 83. Achnanthes glabrata, Grun., 83. subsessilis, Ehrenė., 83. Achras Sapota, Linn., 44. Acicarpha crassifolia, Miers, 486. Actidesmium, Reinsch, 462. Actinoptychus splendens, Ralfs, 85. Actinostrobus, 935, 302, 322. Adesmia muricata, DC., 481. Æcidium elatinum, 284. ZEschynomene, 9, 48. hispida, H. B. K., 3, 29, 71. Agathis, 235, 293-295, 301, 307, 322- 326. Ageratum conyzoides, Linn., 14, 42, 45. Albz (Salices), 337. Albizzia Lebbek, Benth., 34. Alga, Life-history of a Stipitate Fresh- water, by G. Massee, 457. Algæ, 463, 523. Alge of Fernando Noronha, 15, 18; by G. M. Murray, 75. Aloe, 64. Altingia, 252. * Alvelose" 56. ` DT Amansia Duperreyl, Ag., ii- E of ernie Noronha, Keteleeria, 254, Asiocarpa, 254. Lowiana, 254, 272. magnifica, 254, 272. nobilis, 236, 249, 254, 272, 309. rdmanniana, 246 248, 254,079, irr mote numidica, 250, 254. pectinata, Willk., 236, 240, ftnote 241, 249, 249, 95], 954. Pichta, 947, 984, ` Pindrow, 249, 254. Pinsapo, 231, 936, ftnote 241, 243, REŽ% 254-257, 272, 297. ginz-Amalis, 249, 285. achalinensis, 236, 237, 240, 273, sibirica, 236, 237, 240, 249, 254. Subalpina, 263, 286. 55. à Amaranthus, 4, 48. caudatus, Linn., 55. 2P2 540 Amaranthus gracilis, Desf., 55. viridis, Linn., 55. Amaryllides of Patagonia, 499. Ammannia latifolia, Linn., 18, 35. Ampelidez of Fernando Noronha, 26. Amphisbeena, 4, 8, 17. Amphora acutiuscula, Kwetz., 82. lineolata, Kuetz., 82. marina, W. Sm., 82. Amygdalinz, 337. Angeardincese of Fernando Noronha, Anacardium occidentale, Linn., 5, 11, 27 Ancylus, 16. Andrena, sp., 25. Andropogon Schenanthus, Linn., 72. Anona squamosa, Linn., 22. Anonacez of Fernando Noronha, 22. Anthephora, 9. elegans, Schreb., 72. Antheridia, 466. Apetale of Fernando Noronha, 54. Apiocystis, Nag., 462, 525, 526. Aplopappus sericeus, Hook. d Arn., Apocynacese of Fernando Noronha, 45. Apodes (Hymenochate), 97. Arachnoidiseus Ehrenbergii, var. cali- fornica, A. Schmidt, 85. Araucaria, 230, 232, 235, 230, 252, 206-307, 322, 326. Balansee, 250, Bidwillii, 232, 233, 250. brasiliensis, 250, 258, 295. Cookii, 234, 250, 262. Cunninghami, 234, 250. — glauca, 234, excelsa, Laml., 234, finote 242, 250, 257. imbricata, 232, 233, 234, 243, 247, 250-252, 258, 282, 311. montana, 250. Muelleri, 250. Pindrow, 286. i Rulei, 250. ntese (Salices), 337. Aristolochia SCH 507. Aristotelia Maqui, L’ Hérir., 499. Arjona patagonica, Hombr. §- Jacquinot, Artocarpus incisa, Forst., 63. Arundo Donax, Lim., 74. Asclepiadezm of Fernando Noronha, 44 ; of Patagonia, 494, Asclepias linifolia, Decne., 495. mellodora, St.-Hil., 495. Asperococcus intricatus, J. Ag., 77. Aspidium coriaceum, Sw., 500 , le Aspilia Ramagii, Ridley, 5, 42. INDEX. Aster sericeus, Less., 487. Athelia ochracea, Pers., 139. pellicula, Chev., 129. scirpina, 137. Typhe, Pers., 137. Athrotaxis, 235, 249, 296, 297, 303, 304, 322. cupressoides, 250. Gunniana, 249. laxifolia, 250. selaginoides, 250. Atriplex cristata, Afog., var. ?, 498. Pamparum, Griseb., 498. Aucuba japonica, 528. _ Auliscus celatus, var. strigillata, 4. Schmidt, 85. Auricularia, 119, 189. persistens, Sow., 186. phosphorea, Sow., 151. reflexa, Bull., 181, 186. tabacina, Sow., 112. Avicennia, 11. . Ayurana (Salix) Humboldtiana, Mart., ftnote 214. - Azara microphylla, Hook., 477. Baccharis artemisioides, Hook. § Arn., 487. Gilliesii, 4. Gray, 488. glutinosa, Pers., 488. magellanica, Pers., 487. notosergilla, Griseb., 488. serrulata, Pers., 488. valdiviana, Phil., 489. . Ball, the late John, Further Contribu- tions to the Flora of Patagonia, 471. Ballia callitricha, 4g., 501, 506, 509, 514, 526. Ballota suaveolens, Linn., 53. Balsamea, 196. Bangiacese, 464, 468. . “ Bara" of Fernando Noronha, 18. Basella alba, Linn., 11, 56. Basidiomycetes, 80. . Batatas paniculata, Choisy, 47. Bauhinia forficata, Link, 33. Bauhiniese, 33. Berberides of Patagonia, 475. Berberis heterophylla, Juss., 475. Biddulphia Baileyi, W. Sm., 84 mobiliensis, Grun., 84. ulchella, Gray, 84. Biddulphiew of Fernando Noronha, 84, 85 Bignonia, 5. roseo-alba, Ridley, 52. Bignoniacesws of Fernando Noronha, 52. Biota, 302. Bixacese of Patagonia, 477. Blainvillea rhomboidea, Cass., 41. INDEX. Blatta americana, 16. rhaavia, 9. diffusa, Sw., 54. hirsuta, Willd., 54. paniculata, Rich., 54. Boopis, 486. anthemoides, 486. crassifolia, 4. Gray, 486. oragineæ of Fernando Noronha, 45. Borreria parviflora, W. Mey., 40. Botany of Fernando Noronha, On the, by H. N. Ridley, 1. Brachyclados lycioides, Gill., 492. Brachypodium sylvaticum, 71. Brassica alba, Boiss., 22. oleracea, Linn., 92. Bromeliaceæ, quite absent from Fer- nando Noronha, 14. Bromus unioloides, Kunth, 500. Bryonia, 524. racemosa, Xw., 37. Tujuja, Vell., 36. Bryopsis pennata, Lame., 78. Buchu,” 36, Buddleia globosa, Lam., 494. Bulimus Ridleyi, 17. Bumelia, 11, 12. fragrans, Ridley, 5, 43. obtusifolia, 44. Bupleurum fruticosum, 153. Burra,” 60. Bystropogon suaveolens, L' Hérit., 58. oabacincha,” 36. actaceæ of Fernando Noronha, 39; of Patagonia, 485. actus quadrangularis, Webster, 39. Casalpinia Gilliesii, Benth., 481. Ussalpiniere, 34. e Caja,” 27. Janus indicus, Spreng., 94. Calceolaria biflora Let 496. Plantaginea, Sm., 496. Callitris, 235, 236, 263, 264, 296-307, 322, 323. quadrivalvis, 289, Calloria epipora, 125. allus, nvestigation into the True ature of, Spencer Moore, 501-526. ‘alonyction grandiflorum, Choisy, 46. speciosum, var. muricatum, DC., 46. Calycera, 486. Ca ycerem of Patagonia, 486. C ymperes Richardi, C. Muell., 74. anavalia, 10, 63. obtusifolia, DC., 3, 6, 10, 33. c ançançao,” 58, b. antharellus partitus, Berk., 169. apim de Planta," 70. e 541 “ Capinche," 71. Capparidee of Fernando Noronha, 20. Capparis, 11. Cynophallophora, Linn., 4, 5, 20. flexuosa, Vell., 21. frondosa, Jacg., 5, 21. Capres, 337, 338, 401, 450. Capsicum frutescens, Willd., 11, 48. * Carapicho,” 71. Cardiospermum Halicacabum, Linn., Carex, 137, 141. Carica Papaya, Linn., 11, 35. Carnauba Palm, 64. Cassia, 13, 23. aphylla, Cav., 482. falcata, 18. occidentalis, Linn., 18, 34. Tora, Linn., 34. Castanea sativa, Mill., 107. Catasetum, On the Sexual Forms of, with special reference to the Re- searches of Darwin and others, by R. A. Rolfe, 206. Catasetum, 206-225. atratura, Lindl., 216, 222. barbatum, Lindl., 207, 208, 217, 218, 224. Bungerothii, N. E. Br., 220. calceolatum, Lem., 224. callosum, Lindl., ftnote 214, 217, 224. cassideum, Linden § Heichb. f., 225. cernuum, Reichb, f., 217, 218, 224. Christyanum, Reichb. f., 223. Claveringii, Lindl., 219. cristatum, Lindl., 209, 210, 211, 217, 218, 224. ——, var. monstrosum, Hook., ftnote 208, 218. Darwinianum, Rolfe, 218, 224. deltoideum, Lindl., 219, 224. discolor, Lind!., 225. floribundum, Hook., 219. fuliginosum, Rolfe, 218. Garnettianum, Rolfe, 224. laucoglossum, Reichd. f., 224. nomus, Lindl, d* Reich., 207, 214, 219, 222. heteranthum, Rodr., 214, 219. integerrimum, Hook., 228. laminatum, Lindl., 223. longifolium, Lindi., 225. luridum, Lindi., 224. macrocarpum, Kunth, 219. needed Rich, 216, 219, 222. macroglossum, Reich, f., 223. 542 Catasetum maculatum, Kunth, 223. Naso, Lindi., 211, 220, 222. CErstedii, Reichb. f., 220, 223. ———, var. aureum, 221. , var. Pottsianum, 221. pileatum, Reichh. f., 220, 223, planiceps, Lindl., 223. purum, Nees, 223. recurvatum, Link, 223. Regnellii, Rodr., 221. roseo-album, Hook., 225. roseum, Reichb. f., 224. Russellianum, Honk., 224. saccatum, ftnote 214, 223. sanguineum, Lindl., 221, 224. Seurra, Reichb. f., 994. semiapertum, Hook., 223. serratum, Lindl., 223. tabulare, Lindl., 223. thylaciochilum, Lem., 224. tridentatum, Hook., 206-220. trifidum, Hook., 217. Trulla, Lindl., 294. umbrosum, Rodr., 221. variabile, Rodr., 222. violascens, Reichb. f. & Warscew., 24. viridiflavum, Hook., 223. Wailesii, Hook., 223, Warscewiezii, Lindl., 224. ‘Caulerpa, 77, 78. clavifera, 4g., 79. cupressoides, Ag., var. alternifolia, Crouan, mexicana, Sond., 79. prolifera, Lamz., 79. taxifolia, 4g., 79. Cayaponia, 11, 12. Momordica, 36. racemosa, Cogn., 36, 37. Tajuga, Cogn., 15, 36. Cedrus, 236-322. atlantica, 231, 232, 236, 240, 249, 57. Deodara, 231, 236, 249. Libani, Rich., 236, 240, ftnote 242, Cellularia cyathiformis, S ., 190. Cembra, 981 formis, Sow., 190 Cenchrus, 9. echinatus, Linn., 6, 71, 81. levigatus, Trin., 72. pungens, H. B. K., 71. viridis, Spreng., 3, 71. Centroceras clavulatum, J. Ag., 15. reene 232, 235, 249, 296, 297, Fortunei, 240, 241, 244, 311. pedunculata, 324. ——, Var. fastigiata, 245. INDEX. Ceramia, 91. Ceramium rubrum, 85. Ceratophyllum, ftnote 536. Ceratosanthes, 11. angustiloba, Ridley, 37. cuneata, Ridley, 37. Hilariana, 37. rupicola, Ridley, 38. trifoliolata, Cogn., 37, 38. Cercostylos brasiliensis, Less., 489. scabiosoides, Benth. § Hook. f., 439. Cereus, 11. insularis, Hemsl., 3, 4, 12, 39. leucanthus, Pfeiff., 485. Chahreea multifida, DC., 498. rosea, DC., 492. Cheetocerotid of Fernando Noronha, 81, 84. Cheetomorpha antennina, Kuetz., 80. Cheetophora, 523. Cheetophoracez, 526. Chamecyparis, 229, 254-257, 266, 302, 304, 322. sphzeroidea, 288, 302. Chameedoris annulata, Mont., 79. Chara, 523. - Characee of Fernando Noronha, 74, 463. Characium, 4. Braun, 462, 526. Chenopodiacem of Fernando Noronha, 56; of Patagonia, 498. ` ME Chenopodium anthelminticum, Linn., 56. Chloris, 9. barbata, Sw., 73. virgata, Sw., 73. Chlorophyce of Fernando Noronha, 78, 81, 463. Chondrus crispus, 523. Chromosporium, 142, 153. Chrysymenia enteromorpha, Harv.?, Chuquiraga erinacea, D. Don, 491. Cinerascentes (Salices), 337. Cionandra racemosa, Griseb., 97. Cissus sicyoides, Linn., 26. Citrullus vulgaris, Schrad., 36. Cladoderris, 164. Cladophora, 80, 458, 523. minuta, Dickie, 79. Morrisis, Harv., 80. subvaricosa, Dickie, 79. Claviceps purpurea, Tul., 81. Clematis, 475. bonariensis, DC., 474, 475. campestris, St.-Hil., 475. dioica, Linn., 475. Flammula, Linn., 475. Hilarii, Spreng., 474, 475. mendocina, Phil., 415 INDEX, 543 Clematis montevidensis, Spreng., 474. Corticium bicolor, Peck, 157. recta, Linn., 475. Boltonii, Fr., 123. triloba, St.-Hil., 474. : Bupleuri, Roum., 153. Cleome difusa, DC., 21. byssinum, P. Karst., 183. monandra, DC., 91. cærulescens, P. Karst., 153. pungens, Willd., 21. cæruleum, Fr., 151. calceum, Fr., 127, 131, 132, 139, 140, 156, 205. calotrichum, P. Karst., 182. . Carlylei, Massee, 148. Coceoneis scutellum, Ehrenb., 83. carneum, Berk. & Cooke, 141. occonema, sp., 82, caulium, Berk. & Curt., 141. "ëm nucifera, Linn., 64. ceraceum, Berk. d* Rav., 150, 151. chlorinum, Berk. d^ Curt., 154. Spinosa, Linn., 21. Cobæa, 524. Cocooneideæ of Fernando Noronha, Codiolum, A. Braun, 462 Codium tomentosum, 4g., 18. ciliatum, Fr., 1 Ceplastrum, Nig., 462. cinctulum, Quelet, 141. Columbea, 232, 933. cinereum, Sace., 115, 125. "pbretaceu of Fernando Noronha, See kf Curt IAT . eitrinellum, . vy MH. Combretum. 5 12 colliculosum, Berk. § Curt., 134. | » 9, 12, 27, 28. 155 156. 205. C rupicolum, Ridley, 27. comedens, Fr., Gegen "AP qnte op Fernando Noronha, 20, confluens, P Ch 124 ^ 42; of Patagonia, 486. GON, oe Coniferæ, Review of some Points in the convolvens, 8 SR 133. Com arative Morphology, Anatomy, corrugatum, B i Curt.. 133. and , ife: History of the, by M. T. tg 7 ir: 8 asters, 22 cri 1" alen Tou ` C . , ° . M t ; S. T 124. GC SE Ce 18 a aah tte of Fernando Noronha, diaphanum, EW ; Of Patagonia, 495. e | ^, 131. Convolvulus Batatas, Linn., 46. dacolorans B. Iv hl 18. Schi Vins, 47. der Ber 7 Cwi, Li a, Schiecht., 46, . d Copernicia cerifera, Mast., 64. Dregeanam, ier i fz ZE Corallina ceratoides, Kuetz., 77. D engel Ellis, 150. Corden globosa, H B. K., 45. ef done Cooke d Ellis, 142. " Coronha Of, det? "e emplastum, Berk. d Ze m9 risti, . : ei , Corticium Fries, 95, 117, 125, 152, 153, epichlorum, Berk. § Curt. 7, 158, 184 , f, 158, . Fr., 118, 205. acerinum, 130, Peabo P. Karst., 131. albido-carneum, Massee, 142. flame tosum Berk. § Curt., 154. alliaceum, Quelet, 134. a n ans Fr. 142. — 7^ var. aceris, Schulz, 134. flaveolum, Massee, 150. alopecinum, Berk. d Broome, flavescens, Massee, 149. . dum, Berk. & Curt, 149. alutarium, Berk. § Curt., 197. Barium, den mk. § Broome, ambiens, Berk. 4- Broome, 135. "MT ` antarcticum, Speg., 193. d Massee, 154. anthochroum, Fr., 141. fato Taura, Fr., 120. . arachnoideum, Berk., 117, 129, 135, fotidum, Berk. § Broome, 131, 49. i ^ 205. Archeri, Berk., 135. "sii 153. atro-virens, Fr., 155. Wësch uem., 152. Auberianum, Mont., 135. fumosum, Fr:, 1 &uriforme, Berk. & Curt., 157. ;ganteum, Fr., 124, 147. urora, Berk., 141. EE rum, Berk. d Curt., 142. labrum, Berkeleyanum, Ces., 156. el velloides, Massee, 153. -Berkeleyi, Cooke, 133. 544 Corticium hepaticum, Berk. § Curt., byporhnoideum, Berk. 4 Curt., bypopyrehinum, Berk. 4 Curt., incarnatum, 140. interruptum, Berk., 138. isabellinum, Fr., 149. jaganicum, Speg., 149. juniperinum, Fr., 124. Kalchbrenneri, Massee, 143. lactescens, Berk., 130, 138. lacteum, Fr., 132, 139. lacunosum, Berk. & Broome, 138. latitans, P. Karst., 156. lepra, Massee, 130. Leprieurii, Mont., 143. Leveillianum, Berk. & Curt., 197. levissimum, Massee, 132. lilaceum, Rabenk., 156. lilooino-fusoum, Berk. & Curt., Liquidambris, Berk., 148. livido-cæruleum, P. Karst., 152. lividum, Berk., 119, 131. lividum, Pers., 152. Lycii, Cooke, 122. majusculum, .. 125. Marchaudii, Ss 143. martianum, Berk. & Curt., 144. microscopicum, Speg., 124. miniatum, Cooke, 140. modestum, Berk. & Broome, 116. molle, Berk., 114. molle, Berk. & Curt., 150, 151. molle, Fr., 143. Mougeotii, Fr., 112. murinum, Berk. & Curt., 110. muscigenum, Berk. d Broome, 120. myxosporum, P. Karst., 130. nigrescens, Fr., 155. nitidulum, P. Karst., 138. Nyssa, Berk. § Curt., 120. ochroleucum, Berk. § Cooke, 119. ochroleucum, Fr , 184. ochthodes, Berk. & Curt., 144. pactolinum, Cooke d Harkn., 149. pallescens, Massee, 199, - pannosum, Fr., 125, Polis reulum, Berk. & Curt., 121, pellicula, P. Karst., 129, perg datum, Berk. 4 Broome, Petersii, Berk. A Curt., 145. plumbeum, Fr., 159. polygonium, Fr., 139, 144, 199. polygonoides, P. Karst. , 139. polyporideum, Berk, d Curt., 130. INDEX. Corticium populinum, Fr., 121. porosum, Berk. § Curt., 121, 138. portentosum, Berk. d Curt., 129. prasinum, Berk. & Curt., 153. pulchellum, Speg., 125. punctulatum, Cooke, 129. radians, 132. radiatum, 132. radicale, Berk., 187. radiosuin, 131, 133, 139. reflecum, Sacc., 115. resupinatum, Sacc., 115. reticulatum, Berk. & Broome, 154. rhabarbarinum, Berk. & Broome, 115. rimosissimum, Berk. d Curt., 122. roridum, Speg., 128. rosellum, Speg., 145. roseolum, Massee, 140, 205. roseum, 140. rubropallens, Massee, 145. rude, P. Karst., 128. salicinum, Fr., 118, 205. salmonicolor, Berk. & Broome, 122. Sambuci, Fr., 137. sanguineum, Fr., 146. sarcoides, Fr., 122. scariosum, Berk. § Broome, 123. scirpinum, Wint., 136. scutellare, Berk. § Curt., 128, 142. seraceum, Fr., 127. seriale, Fr., 126. seriatum, Berk., 204. serum, Fr., 127. simulans, Berk. & Rav., 119, 120, 128 siparium, Berk. § Curt., 139. sordidum, 140. subalutaceum, P. Karst., 139. subcontinuum, Berk. & Curt., 128. subrepandum, Berk. 4 Cooke, 119, 14 subsulphureum, P. Karst., 148. suffultum, Berk. § Broome, 158. tenuissimum, Berk. § Broome, 123. terreum, Berk., 158. tessulatum, Cooke, 136. tremellinum, Berk. 4. Rav., 146. ——, var. reticulatum, Massee, 146. tritiusculum, Berk. & Broome, 111. triviale, Speg., 126. Typhs, Fuckel, 137, 141. Ulmi, Lasch., 157. uvidum, Fr., 156. vagum, Berk. & Curt., 133, 148. vellereum, Ellis 4. Cragin, 137 velutinum, Berk., 143. venosum, Berk. 4: Rav., 147. versiforme, Fr., 126. violaceo-lividum, Fr., 151, 152. INDEX. Corticium viride, Preuss., 157. viscosum, Berk., 119. viticola, Fr., 146. Cosrinodiscea: of Fernando Noronha, , 85. Coscinodiscus anguste-lineatus, A. Schmidt, 85, argus, Ehrend., 86. asteromphalus, Ehrend., 86. biangulatus, 4, Schmidt, 85. denarius, 4. Schmidt, 85. heteroporus, Ehrenb., 86. Intumescens, Pant., 86. marginatus, Ehrenb., 86. minor, Ehrenb., 85, Oculus-iridis, Ehrenb., 86. perforatus, Ehrend., 86. radiatus, var. media, Grun., 86. robusta, Grev., 86. Cosmocladium, Bréb., 462. terellus cornucopioides, 164. Tessa cretica, Linn., 495. Cristaria, 478, Crotalaria, 23, striata, DC., 28. Croton odora tus, Ridley, 11, 58. populifolius, 59. ucibulum vulgare, 168. ruciferæ of Fernando Noronha, 22; of Patagonia, 476. Oryptogamia of Fernando Noronha, 74. Oryptomeria, 235, 230, 241, 250, 304, elegans, 262. poen, 236, 241, 257, 262, 295, Lobbii, 237, 240, 262. torta, 262. Cucumis, 524. nguria, Linn., 47. Melo, Linn., 36. Cucurbita, 504. Pepo, Linn., 36, 502. Oucurbitaceæ of Fernando Noronha, 11, 32, 13, 15, 21, 36; of Patagonia, Cunninghamia, 235, 236, 249, 270, 292- sinensis, 270, 305. Cupressinese, 235, 292-328. Oupressus, 235, 236, 238, 250, 288, 298-807, 322-326. funebris, 240, 257, ftnote 265. Cupressus pyramidalis, 236, retrofracta, 236, sempervirens, 240, 241, 257. Cupuliferee of Patagonia, 499, Cuscuta americana, Linn., 47. decora, 48. lobosa, Ridley, 48. Cutléria, 469. D Cutleriacese, 463, 467, 469. Cyanophyoes, 80. Cyathus striatus, Hoffm., 81. Cycas, 300, 301, 323. Cymbella amphicephala, Naeg., 82. obtusa, Greg., 82. Cymbelleæ of Fernando Noronha, 82. Cynosurus indicus, Linn., 72. Cyparissidium, 263. Cyperaceg of Fernando Noronha, 64, Cyperus, 65. atlanticus, Hemsl., 11. atlanticus, Hemsl., 66. brunneus, Sw., 3, 11, 66. circinatus, Ridley, 64, 95. compressus, Presl, 65. distans, Linn., 67. ferax, Rick., 67. insignis, Kunth, 66. ligularis, Linn., 3, 66, 67. Noronhz, Ridley, 66. olidus, Rich., 66. planifolius, Rich., 66. purpurascens, Vahl, 66. rotundus, 65. vialis, Ridley, 65. viridis, 65. Dacrydium, 235, 250, 261, 292-328, Colensoi, 261. elatum, 261. Franklinii, 299. Dactylena micrantha, Schrad., 4, 21. Dactylococcus, Nag., 462. | Dactyloctenium egyptiacum, Willd., 72. Dammara, 307. Dandelion, 528. Datura fastuosum, Linn., 49. Stramonium, Linn., 12, 49. Davies, Thos., Geology of Fernando Noronha, 86. Desmodium, 9. barbatum, Benth., 30. incanum, DC., spirale, DC., 30. triflorum, DC., 29. Diandre (Salices), 397, 338, 344. Diatoma flabellatum, Juerg., 83. Diatomaces of Fernando Noronha, by J. Rattray, 81, 462, 463. ichromena micrantha, Kunth, 67, SC feroz, Schrad., 67. 546 .INDEX. Dictyopteris, Lam., 463-410. Escallonia, 483. polypodioides, 464, 466, 467, 469. alpina, Poepp., 484. Dictyosphseria, 526. corymbosa, Pers., 484. Dictyospherium, Nág., 457—462. Ehrenbergianum, Nág., 457, 458, 462. Hitcheockii, Wolle, 458. pulchellum, Wood, 458. reniforme, Buln., 457. Dictyota, 464—468. bartyresiana, Lam., 77. ciliata, J. 4g., TT. dichotoma, Huds., 77, 464, 466, 467 Dietyotacese, On the Systematic Position of the, with special reference to the Genus Dictyopteris, Lam., by Thos. Johnson, 463-470. Digitaria horizontalis, Willd.. 69. Dimorphococcus, A. Braun, 462. Diplopappus coronopifolius, Less., 487. Diselma, 302. Doldinia concentrica, 81. Dolichos obtusifolius, Lam., 33. Dombeya excelsa, 242. Dracena, 465. Drosera, 532. Duvaya longifolia, 125. Echinocactus leucanthus, Gill, 485. Echinonyotanthus leucanthus, Lemaire, A Ecirrosse (Catasetum), 224. Ecklonia baccata, 523. Eclipta alba, Hassk., 42. erecta, Linn., 42. Ectocarpus, 469. fulvescens, Thuret, ftnote 469. pusillus, 469, 470. secundus, Crouan, 469, 470. siliculosus, 469, 470. Zebelii, Kuetz., 469, 470. Elainea Ridleyana, Sharpe, 14, 44. Eleocharis, 13. Eleusine ægyptiaca, 5. cruciata, Lam., 72. indica, Gaertn., 72. Elvella lilacina, Batsch, 186. pannosa, Sow., 164. Embothrium coccineum, Forst., var. ?, lanceolatum, Ruiz 4^ Pav., 498. Enteromorpha compressa, Grev., 79. Equisetum, 296, 297. Eragrostis ciliaris, Link, 9, 73. Erythrina, 10, 32, ftnote 208. aurantiaca, Ridley, 30, 95. exaltata, Webster, 14, 18, 31. Kos, Willd., 32. ulungu, Mart., 31, 32. macrantha, 532. Philippiana, Engl., 484. stricta, C. Gay, 484. Escobilla, 488. “ Espongeira," 35. Estrella de la vispera, 484. Eucalyptus, 116. Eueatasetum, 215, 222, 223. Euphorbia capitata, Lam., 56. Chammsyce, 499. comosa, Vell., 11, 56. cuspidata, Bertol., 57. Engelmanni, Boiss., 499. hirta, Linn., 56. hypericifolia, Linn., 11, 57. pilulifera, Linn., 56. prostrata, Ait., 499. thymifolia, Burm. f. 9T. ,var. B, Linn., 57. Euphorbiacese of Fernando Noronha, 11, 56; of Patagonia, 499. Eupodisces of Fernando Noronha, 85. Euxolus caudatus, Moq., 55. viridis, Moq., 9, 55. Evania, 16. Exidia glandulosa, Fr., 80. Fabiana imbricata, Ruiz § Pav., 496. Fagus antarctica, 129, 191. obliqua, Mirbel, 499. procera, Poepp. § Endl., 499. ** Fedegozi,” 45. “ Feijao de lenha,” 20. ` Fernando Noronha, Notes on the Botany of, by H. N. Ridley, 1. Ficoideæ of Fernando Noronha, 38. Ficus, 11. . Noronha, Oliv., 3, 62. Filices of Patagonia, 500. Fimbristylis diphylla, Vahl, 67. glomerata, 11. Fitzroya, 235, 250, 302. patagonica, 256, 302. Fleurya æstuans, Gaudich., 63. . Flora of Patagonia, Further Contri- tions to the, by the late John Ball, 471. Flora, Relations of, to Insect Fauna, Fernando Noronha (Ridley), 12. Florideæ, 75, 463—470. ep Fragilaries of Fernando Noronha, 85. Fragiles (Salices), 337, 338, 362-364. Fregata aquila, 4. Frenela, "235, 540, 241, 250, 261-270, 293-328 “ Frutta di Sapo,” 45. | Fucaceæ, 463, 465, 467. INDEX. Fucus, 469. vesiculosus, 465. Fungi-of Fernando N oronha, by G. M. urray, 80. Furerga gigantea, 64. Gaillardia scabiosoides, Benth. 4- Hook. Galaxaura, 81. Cylindrica, Lamz., 76. lapidescens, Sol., 76. oblongata, Lamz., 76. rugosa, Lame., 76. Galegeie, 28. “ Gamaleira," 63, Gammarus, 15, 16. Gamopetale of Fernando Noronha, 40. Gelidium cartilagineum, 523. Genistex, 28, Gentianes of Fernando N. oronha, 45. Geology of Fernando N oronha, by Thos. Davies, 86. "o, A., Musei of Fernando Noronha, Geraniacess of Fernando Noronba, 25; of Patagonia, 480. Geranium intermedium, Bet., 480. . Patagonicum, Hook., 480. Gigartina Teedii, Lam., 75. Ginkgo, 232-235, 249, 258, 275, 292- 300, 323, . biloba, 240, ftnote 242, iraudia sphacelurioides, 469. Gito, 21. Glycyrrhiza astragalina, Gill., 480. Glyptostrobus, 263, 285. naphalium americanum, Zinn., 489. purpureum, Linn., 489. Spicatum, Lam., 489. Gomphonema paradoxum, C. 4g., 83. Gonolobus, 10. micranthus, Hemsl., 14, 44. ossypium barbadense, Linn., 22. Gracilaria armata, Ag. ?, 76. multipartita, Clem., 76. Grahamia bracteata, Gill., 477. ramineg of Fernando Noronha, 68; . of Patagonia, 500. ‘Gramma,” 69. Grammatophora angulosa, var. hamu- lifera, Grun., 84. Japonica, Grun., 84. ——— marina, var. intermedia, Grun., 84. Grevillea robusta, 528, 530, 539. Grindelia diffusa, Gill., 486. pulchella, Dun., 486. speciosa, Gil/., 487. Guadua latifolia, Kunth, 74. Guettarda, Il. t ‘ae `a 547 Guettarda lica, 40. Leai, Ridley, 40. viburnoides, 40. Guntheria megapotamica, S; ., 489. Gymnopogon rupestre, Ridley, 3. Halictus, 13. . Halimeda Opuntia, Lamz., 78, Halyseris, 4g., 463. delicatula, Lame., 48. Justii, Lamz., 78. plagiogramma, Mont., 78. Haplopappus coronopifolius, DC., 487. Haplospora globosa, Kjellm., 469. edysares, 29. Hedysarum barbatum, Linn., 30. Helichrysum, 486. Heliopeltez of Fernando Noronha, 85. Heliophytum indicum, DC., 45. Heliotropium indicum, Linn., 45. Helvella nicotiana, Bolt., 112. Hemicarpha isolepis, 67. Hemidactylus mabouia, 16. Hepatic of Fernando Noronha, 74. Hibiscus esculentus, Linn., 22. Hippeastrum, 500. Hoffmannseggia falcaria, Cav., 482. trifoliata, Cav., 482. Holcus halepensis, Linn., 72. Homoianthus echinulatus, Cass., 493. Homotaxy, 243. Hormospora pellucida, Dickie, 80. Hyalis argentea, D. Don, 491. Hydnum, 117. aurantiacum, Berk., 180. Hydradephaga, 16. Hydrianum, Rab., 462, 526. Hydrocytium, A. Braun, 462. Hylobolus tumulosus, P. Karst., 204. Hymenocallis, 64. Hymenocheete, 106, 111, 114, 117, 128, 159, 196, 203. abietina, Massee, 115. agglutinans, Ellis, 106. ambiens, Berk. d Curt., 106. aspera, Berk. 4 Curt., 100. attenuata, Lév., 98. avellana, Cooke, 103. badio-ferrugiuea, Lév., 101. barbata, Massee 109. Berkeleyana, Cooke, 101. Boltoni, Cooke, 114. hohserrettts 1 ., 108. Cacao, Berk., 100. cervina, Berk. d Curt., 114. ia, Berk. & Curt., 196. ata, Lév., 110. sorticelor , Berk. § Rav., 111, 112. crassa, Berk., 114. crocata, Lév., 105. 548 INDEX. Hymenochsete croceo-ferruginea, Massee, 110, 205. crocicreas, Berk. § Broome, 107. damecornis, Lév., 96. depallens, Berk. § Curt., 107. Dregeana, Massee, 114. dura, Berk. § Curt., 105, 111. elegantissima, Massee, 99. episphzeria, Massee, 111. ferruginea, Massee, 103. fimbriata, Ellis §- Everh., 113, 205. flavum, Berk., 117. formosa, Lév., 96. fuliginosa, Lév., 109. fulvella, Berk. § Curt., 104. imbricatula, Lév., 103. innata, Cooke d^ Massee, 109. insularis, Berk., 107. Kalchbrenneri, Massee, 116. Kunzeii, Massee, 100, 173. leta, Berk., 98, 99, 185. levigata, Massee, 107. leonina, Berk. & Curt., 107. modesta, Massee, 116. Mougeotti, Massee, 111, 205. multispinulosa, Peck, 108. nigrescens, Cooke, 104, 205. olivacea, Cooke, 116. pallida, Cooke / Massee, 97. . paupercula, Berk. & Curt., 121, pellicula, Berk. 4^ Broome,106, 205. perpusilla, Pat., 104. phæa, Berk., 98. pulcherrima, Massee, 104, 205. purpurea, Cooke et Morg., 115. ramealis, Berk., 187. reniformis, Lév., 96, 205. rhabarbarina, Massee, 113. rheicolor, Lév., 98. rheicolor, Mont., 185. rigidula, Berk. § Curt., 99. rubiginosa, Lév., 97, 103, 105, 106. Sallei, Berk. & Curt., 101. scabriseta, Cooke, 113, 205. Schomburgkii, Massee, 115. siparia, Berk. & Curt., 204. spadicea, Berk. § Broome, 102. speciosa, Lév., 97. spreta, Peck, 108. stelligera, Speg., 110. Stevensoni, Berk. & Broome, 106. strigosa, Berk. & Broome, 102. subpurpurascens, Massee, 101. tabacina, Lév., 99, 101, 112, 173, tasmanica, Massee, 105, 205. tenuis, Peck, 109. tenuissima, Berk., 102. toxia, Berk., 108. tristiuseula, Massee, 111, Hymenochætə tuberculosa, Cooke, 112. umbrina, Massee, 113, 116. unicolor, Berk. § Curt., 108. veluticeps, Berk., 116. vibrans, Massee, 117. Hypheotrichis, 145. Hypnunn, sp., 74. Hypochnus, Fries, 117, 132. flavescens, Bon., 149. fumosus, Fr., 153. Hyptis pectinata, Poit., 53. suaveolens, Poit., 53, 73. Ianthina, 10. Incanæ (Salices), 337. Indigofera, 18. Anil, Linn., 28. Insect Fauna, Relations of, to the Flora of Fernando Noronha (Ridley), 12. Ipomea Batatas, Lam., 3, 46. digitata, Linn., 47. grandiflora, Lam., 46. muricata, A7: 46. 10, 47 entaphylla, Jacg., 3, 10, 4/. E ice, Sweet, 5, 6, 10, 46, 48. Quamoclit, Linn., 46. Tuba, Don., 5, 6, 10-12, 17, 46. Tresine aggregata, Moq., 55, 56. vermicularis, Moq., 55. Isoétes, 263, 298, 319. Jacquemontia euricola, Ridley, 47. Jacquinia, 11. armillaris, Jacg., 5, 24, 43. “Jajo,” 19. ` Jambosa vulgaris, DC., 35. Jania cubensis, Mont., 77. rubens, Lamz., 77. Jarilla, 479. Jatropha, 18. Curcag, Linn., 57. genuina, vars, Mucll- Ag. 58. ossypifolia, 57. Pohlinna, Muell.-Arg., 5, 10, 57. —, var. subglabra, 57. urens, Linn., 10, 58. . Johnson, Thos., On the Systematio Position of the Dictyotaces, wit special reference t to the Genus Dicty- opteris, Lam., 463. d uniperus, 229, 235, 249, 250, 299-303, 324. conferta, 264. drupacea, 250. pheenicea, 261. sinensis, 261. taxifolia, 264. Jussieua, 14, 35. acuminata, Sw., 35. linifolia, Vahl, 13, 14, 36. 549 INDEX. Kallonema obscurum Dickie, 79. Loasacer of Patagoni Keteleeria, 236, 293. " Lobaria pulmonatis,. Se Davidiana, 236. Fortunei, 251. Labiate of Fernando Noronha, 53. Lagenaria vulgaris, 502. Laguncularia racemosa, Gaertn., 10, Laminariæ, 519. Lantana amena, Ridley, 52. cinerascens, 53. Jucata, Lindl., 52. acina, Desf., 52. Larix, 296 dde V. europea, Duch., 286, ftnote 242, 295, 312, 313. Griffithii, 236, 257. leptolepis, 285. microcarpa, 236. sibirica, 236, Larrea cuneifolia, Cav., 480. divaricata, Cav., 479. nitida, Cav., 480. Lathyrus pubescens, Hook., 481. tomentosus, Lam., 481. Laurencia papillosa, Forsk., 76. Scoparia, J. 4g., 76. Laurus, 177. Leguminosæ of Fernando Noronba, 9, 2-15, 28, 47-95; of Patagonia, 480. Lepidodendron, 298. Loucoria achilleifolia, Hook. 4 Arn., 93. runcinata, Gill., 492. Leuchæria, 492, 493. Leucopsis sericea, Baker, 487. » var. eriophora, 487. Leuzites erubescens, Berk., 80. Libocedrus, 235, 241, 250, 263, 302, 322, 326. austrocaledonicus, 266. chilensis, 266. decurrens, 257, 266, 281, 287. Doniana, 256. macrolepis, 257. tetragona, 249, 256, 302. . Licmophora australis, Grun., var. major, debilis, Grun., 83. Lyngbyi, var. longa, Grun., 84. paradoxa, 4g., 83. Linaria canadensis, Dum. Cours, 496. Liquidambar, 139, 148, 153. Lithothamnion, 81. mamillare, Harv., 77, polymorphum, 4resch., 77. Loasa filicifolia, Poepp., 485. pinnatifida, Gi., 485. Loganiaceæ of Fernando Noronha, 44; of Patagonia, 494. Longifoliz (Salices), 337. Loranthaces of Patagonia, 498. Loranthus tetrandrus, Ruiz § Pav., 498. Lucidz, 337. Luffa, 36. cylindrica, M. Roem., 36. operculata, Cogn., 36. purgans, Mart., 36. Lycium, 122. . Lycopersicum esculentum, Mill., 11, 48 Lycopodium, 254-257, 263, 298. annotinum, 266. casuarinoides, 264. Selago, 319. serrulatum, 258. spectabile, 247. tetragonum, 256. volubile, 247. Lymnea, 16. . Lyngbya Noronhe, Dickie, 77. Lyomyces, 133. serus, P. Karst., 197. Lythraries of Fernando Noronha, 35. Macrocystis, 502. pyrifera, Ag., a Short Note on (Opona oore), 519, 520, 526. Macrostylæ (Salices), 337. , Majai, 491. Malachra capitata, Linn., 28. radiata, Hemsl., 23. Malva, 478. bonariensis, Cav., 478. crispa, Hook., 479. obtusiloba, SE? 478. tagonica, . Eilphurea, Géi, 418. Malvaceæ of Fernando Noronha, 9, 22 ; of Patagonia, 478. Malvastrum, 478, 479. sulphureum, Griseb., 478. Mangifera indica, Linn., 27. Manihot utilissima, Pohl, 57. Manzana del Diablo, 496. Marchantia, 296, 297. Margarita rosada, 500. a Massee, G., Life-History of a Stipitate Freshwater Alga, 457. ——, Monograph of the Thelephores, " Masters, Maxwell T., Review of some Points in the e Morpho- logy, Anatomy, an Lite-History of the Ooniferse, 226. Melanophycese, 463. 550 Melica rigida, Cav., 500. Melosira nummuloides, 4g., 84. Melosiress of Fernando Noronha, 84. Merulius, 200. Corium, 128. Micantes (Salices), 337. Microcachrys, 235, 293, 296, 297, 301. Microstyls (Salices), 337. Mimosa pudica, 533, 537, 538. Mimoser, 34. Mimulus luteus, Linn., 497. Mischococcus, Nàg., 462. Momordica, 11. Charantia, Linn., 3, 13, 15, 36, Monachanthus, 211-222. Bushnani, Hook., 225. discolor, Lindl., 225. fimbriatus, Gardn., 225. viridis, ftnote 207, 208-220. Monocotyledons of Fernando Noronha, Moore, Spencer le M., Studies in Vege- table Biology.—VI. An Investigation into the True Nature of Callus: The Vegetable-Marrow and Ballia calli- tricha, Ag., 501-526. —. VII. Some Microchemical Reactions of Tannin, with Re- marks upon the Function of that Body, and its Excretion from the General Surface of Plants, 527-538. “ Mororo,"' 33. Morus albus, 179. Mucuna urens, DC., 10, 33. * Mulungu vermeho,” 31, 32. Murena, 18. Murray, G. M., Alge of Fernando Noronha, 75. -—, Fungi of Fernando Noronha, 80. Mus rattus, 8. Musa sapientum, 528. Musei of Fernando Noronha, by A. Gepp, 74. Mutisia ilicifolia, Cav., 491. spinosa, DC., 491. truncata, D. Don, 491. Myanthus, 208-224. barbatus, ftnote 207, 208-210, 214, cernuus, Lindl., 209, 210, 217. cristatus, Lindl., ftnote 211, 218. deltoideus, Lindl., 219. sanguineus, Linden, 221. Myrica, 113. Myrsines of Fernando Noronha, 43. Myrtacex of Fernando Noronha, 35. Nassauvia rosulata, Hook. § Arn., 493. Navicula brasiliensis, Grun., 82. INDEX. Navicula cocconeiformis, Kuetz., 82. erythrea, Grun., 82. interrupta, Kuetz., var., 83. laciniosa, A. Schmidt, 82. minuscula, var. bahusiensis, Grun., 82. mutica, Kuetz., 82. subula, Kuetz., 82. Weissflogii, 4. Schmidt, 82. Navicules of Fernando Noronha, 82. Nepeta pectinata, Linn., 53. Nephrocytium, 526. Nereocystis, 502, 519. Nicolsonia, 30. reptans, Meissn., 29. Nicotiana acuminata, Hook., 496. angustifolia, Ruiz § Pav., 496. tabacum, Linn., 49. Nidularium, 524. ` Nierembergia filicaulis, Lindl., 495. linifolia, Miers, 495. rigida, Miers, 495. Nitella, 15. cernua, A. Br., 6, 74. Nitidulæ, 337, 339, 432. Nitzschia fluminensis, Grun., 84. lanceolata, W. Sm., 84. . marginulata, var. subconstricta, Grun., 84. marina, Grun., 84. Nive, 337, 339, 421. Noticastrum eriophorum, Remy, 487. Nycteginez of Fernando Noronha, 54. Nyssa, 120. Odontia crustosa, 130. papillosa, 130. CEdogonium, 523. CEnothera odorata, Jacq., 484. stricta, Ledeb., 484. Onagrariex of Fernando Noronha, 35; of Patagonia, 484. Oocardium, Naeg., 462. Oocystis, 526. Oogonia, 467. Ophiocytium, Rabenh., 462. Opuntia grata, Phil., 485. Oreodoxa regia, Kunth, 64. “ Ortega branca,” 58. * Ortega trepadeira," 60. Oryza sativa, Linn., 12. Osmunda, 291 Oxalis, 6, 25. Noronhæ, Hook., 4, 5, 12, 13, 24. stricta, 25. - sylvicola, Ridley, 25, 95. valdiviensis, C. Gay, 480. Padina, 466. pavonia, Linn., 77. INDEX, Pepalanthus, 13. Palicourea, 5, 11. 12, 24. insularis, Ridley, 41. Palms of Fernando N oronha, 64. Paludina, 16. Panicum appressum, Lam., 69. brevifolium, 70. brizoides, Lam., 6, 13, 69. capillaceum, Lam., 70. caudatum, Lam., 71. ciliare, Retz., 69. Colonum, Linn., 70. Jasciculatum, N ees, 70. fuscum, Sw., 70. Leandri, Trin., 69. numidianum, Lam., 6, 70. parvifolium, Lam., 70. plantagineum, Link, 69. sanguinale, Linn., var. ciliare, Doell, , var. distans, Doel/, 69. scandens, Trin., 71. trichodes, Sw., 70. Papaw, 12, 35. apayaceæ of Fernando Noronha, 35. Paspalum anemotum, Ridley, 68. brizoides, 35. phonoliticum, Ridley, 6, 68, 95. virgatum, 68. atagonia, Further Contributions to the Flora of, by the late John Ball, Pavonia cancellata, Cav., 5, 22. Pellæa concolor, Bak., 74. , geraniæfolia, Fée, 74. Peltigera canina, 523. Peniophora, 96, 114, 117, 140, 158, 159 cinerea, 111, 152. disciforme, Cooke, 189. gigantea, Fr., 124, 147. incarnata, 140. quercina, 124. rosea, 140. Pentandra (Salices), 337, 338, 359. erdicium recurvatum, Vahl, 498. Perezia recurvata, Less., 493. Pernettya mucronata, Gaudich., 493. Peroniella, Gobi, 469. Petasites vulgaris, 528, 531, 532. Petunia acuminata, R. Grah., 496. Peyssonnelia Dubyi, Crouan, 76. eziza sarcoides, Wahlenb., 118. Phseophycese, 463, 464, 465, 466-470. Phæophyces of Fernando Noronha, Pherosphora, 293. Phaethon sethereus, 4. Phaeeolee, 30. Phaseolus bipunctatus, Jacq., 93. 551 Phaseolus lunatus, Linn., 3, 33. peduncularis, H B. K., 15, 33. Pherosphsra, 235, 299. Philibertia Gilliesii, Hook. d Arn., 494. Philoxerus, 10, 48. aggregatus, H. B. K., 55, 56. longespicata, Seubert, 56. vermicularis, F.. Br., 3, 6, 10, 13, 33, 55, 63. Phragmites communis, ftnote 508, 518, 519. Phylicifolise (Salices), 337, 338, 395, 401. Phyllanthus, 3. brasiliensis, Muell.-Arg., 58. commutata, var, Muell.-Arg., 58 lathyroides, Muell.-Arg., 58. Phyllocladus, 235, 241, 257, 276, 297, 300, 301. Physalias, 10, 18.' Physalis, 11. hirsuta, Dunal, 49. —, var. ?, Hemsl., 49. minima, Linn., 49. viscida, Ridley, 49. Phytolaccacer of Fernando Noronha, 56. Picea, 236, er i 272-328. alba, 250, 313. ajanensis, 239, 243, 248, 249, 272, 274, 284. 272, 973, 974 Engelmanni, . eeng Link, 231, 290, 287, ftnote 242, 255, 257, 274, 282, 283, 295. —, var. denudata, Koch, ftnote , 296, —, var. monstrosa, 245. ——, var. pygmga, 245. ——, var. viminalis, 245, ftnote On virgata, Will, ftnote 283, 316, 318. glauca, 295. ezoensis, E Khutrow, 250. Menziesii, m 239, 240, 241, 243, 257, 272, 284. Morinda, Coe 274. nigra, 250, 272, 313. orientalis, 236, 240, 257, 291. polita, e ga ungens, 273. PE eae. glauca, 250. rubra, 272, 273. sitchensis, 249. amabilis, 292. 552 INDEX. Pinus argentea, 252. aristata, 236, 254. arizonica, 25]. australis, 236, ftnote 242, 295. austriaca, 231, 236, 295. Ayacuite, 236, 251, 254. Balfouriana, 236, 254, 295. Banksiana, 236, 254, 295. bracteata, 254. Bungeana, 236, 272, 274. calabrica, 236, 239. canariensis, 236, 237, 238, 239, 252, 295. Cembra, 236, ftnote 242, 252, 254, 272, 274, 284, 295, 310. cembroides, 236, 254. cephalonica, 236. Chihuahuana, 295. clausa, 254. contorta, 141, 251, 254, 274, 295. Coulteri, 236, 251, 254, 272, 285, 295, 310. cubensis, 254, 295. ——, var. tertrocarpa, 309. deflexa, 252. densiflora, 252, 274, 295. Sulis "n 258, 274, 295. iottii, elm., 236, ftnote 242, 258, 295. Engelmanni, 251, 254. excelsa, 236, 239, ftnote 242, 250, 252, 260, 272, 274, 285, 295. flexilis, 236, 251, 254, 295. Fremontiana, 270, Gerardiana, 236. glabra, 236, 254, 295. Greggi, 295. halepensis, 236, 248, 295. hudsonica, 252, 274, inops, 236, 252, 254, 274, 295. insignis, 231, 236, 252, 254, 258, insularis, 295. Jeffreyi, 236, 240, 252, 254, 274. asia, 295. khasyana, 258. Lambertiana, 236, 252-254, 260, 295. Laricio, 231, 236, 237, ftnote 242, 252, 254, 274, 295, 310. latisquama, 254. . leiophylla, 295. Lemoniana, 309, 320. Llaveana, 274. longifolia, 252. maderensis, 236. maritima, 236, 237, 252. Maesoniana, 236, 274, 295. Merkhusii, 295. Pinus mitis, 236, 252, 254, 274, 295. monophylla, 236, 252, 254, 269, 270, 272, 274, 295. monspeliensis, 236, 239, ftnote 242 montana, 236, 275, 295. Montezume, 251, 254, 295. monticola, 236, 251, 252, 254, 274, 295. muricata, 251, 254, 274, 285, 295, 22 322. Murrayana, 114. nigricans, 252, 295. obovata, ftnote 283. alustris, 236, 237, 254. Parryana, 236, 254, 258, 274, 295. parviflora, 236. patula, 250. Peuke, 236, 252, 274, 295. Pinaster, 236, ftnote 242, 252, 254, 295. Pinea, Rich., 286-239, ftnote 242, 252, 258, 259, 274, 295. Pithyusa, 236. ] ponderosa, 231, 236, 251, 252, 254, 274, 288, 295. Pumilio, Nees, ftnote 242, 274, 295, 310. ` pungens, 236, 254, 274, 299. pyrenaica, 252, 274, 295. recurva, 310. reflexa, 251, 254. religiosa, 254. resinosa, 236, 251, 274, 295. rigida, 231, o ao 254, 258, 274, 285, 295, 313. Sabiniana, 236, 254, 258, 274, 295. Salzmanni, Dun., ftnote 242. serotina, 252, 254, 285. silvestris, Willk., 231, 236, 240, ftnote 242, 252. 254, 258, 269, 272, 274, 284, 295. Strobus, Henry, 236, 237, ftnote 242, 250, 251, 252, 254, ftnote 257, 274, 284, 295, 310. sylvestris, 130, 140, 157, 518. T:eda, 236, 252, 254. Thunbergii, 295, 313, 314. _ Torreyans, 236, 241, 252, 254. "m tuberculata, 236, 251, 254, 274, 295. uncinata, 275, 309. viminalis, A/str., ftnote 283. Pisonia, 10. Darwinii, Hemsl., 5, 54. Planorbis, 6, 15, 16. Plantaginee of Fernando 53; of Patagonia, . " Plantago Bridzesii, Decne., var. angusti folia, Oliv., 498. Noronha, INDEX. Plantago major, Linn., 53. Plants introd. by Sea-currents, Fer- nando Noronha, 10. Pleiandræ (Salices), 337, 338, 344, 347. eurococcus vulgaris, Meneg., 459. Pleurosigma Lorenzii, Grun., 88. speciosum, W. Sm., var., 83. Plocamium, 286. umbagineæ of Fernando Noronha, 43; of Patagonia, 493. Plumbago, 9. occidentalis, Sweet, 43, scandens, Linn., 6, 43. Poa ciliaris, Linn., 13. Odisoma maeropus, 146. Podocarper, 235, 309, 322. Podocarpus, 235, 247, 249, 250, 290, 296, 297, 301, 324. chinensis, ftnote 311. cupressina, 261. dacrydioides, 261, 294. Lamberti, 295. Sellowii, 295. Podosphenia debilis, Kuetz., 83. odostylæ (Salices), 337. Pæpalanthus (err. for Pæpalanthus), 13. Polistes, 13. . Polypetale of Fernando Noronha, 20. olyporus, 197. affinis, 172. fomentarius, Fr., 80, 523. hirsutus, Fr., 80, igniarius, Berk., 107. lucidus, Fr., 80. luteus, 171. sulphureus, 523. Polytrie um, 523. Pompilus nesophila, 13. Porp yra, 465. ortulaca, 11. grandiflora, Hook., 477. oleracea, Linn., 10, 22, 477. patens, Jacq., 22. Portulacacem of Fernando Noronha, 22; of Patagonia, 477. Potentilla Anserina, 528. Prima vera,” 46. Primulaceæ, 312. Prosopis, 483. cinerascens, 483. reptansi, 483. P strombulifera, Benth., 482, 483. roteacess of Patagonia, 498. Pruinosg (Salices), 337. Prumnopitys, 249, Funus Armeniaca, 203. Pseudocatasetum, 207. Pseudolarix, 275, 309. Kaempferi, 251, 253, 310. Pseudotsuga, 236, 309. 553 Pseudotsuga P Douglaeii, 231, 236, 240, 243, 244, 249, 254, 971 297, 309, 316. Hs S08 ——-, var. brevifolia, 241. Psidium Guyava, Raddi, 95. ` Psilopus metallifer, 13. Purpures, 337, 340, 447. Pylaiella fulvescens, Thur., 469, 470. nana, Kjelim., 469. littoralis, 469, 470. Fyrenomyeetes of Fernando Noronha, Pyrus Aucuparia, 156. “ Quichaba," 44. Quinchamalium majus, Brongn., 498, 499. Rangia, 465. Ranunculacee of Patagonia, 474. Rat Island (Ridley), 3. Rattray, J., Diatomacee of Fernando Noronha, 81. Rauwolfia, 11. ternifolia, Kunth, 45. Reichartia rosea, Karsten, 45. Remirea maritima, 11. Repentes (Salices), 337, 338, 389. Resupinate (Hymenochsete), 104. Retamilla, 494. Retinospora, 235, 254-257, 288, 311. leptoclada, 261, 302. obtusa, 237, 240, 241. pisifera, 237, 240, 241. squarrosa, 262, 264. tetragona aurea, 256. * Retorton,” 483. Rhizosolenia, 81. styliformis, Brighiw., 84. Rhodophycee, 81, 84. Rhynchosia minima, DC., 10, 34. Rhynchospora micrantha, Vahi, 67. Riccia limbata, Bisch., 75. Ridleyi, Gepp, 74. Ricinus, 11, 524. . brasiliensis, var., Muell.-Arg., 57. communis, Linn., 3, 10, 57. Ridley, H. N., Notes on the Botany of Fernando Noronha, 1. Rigidse (Salices), 337. Rivina, 11. levis, Linn., 56. Rolfe, R. A., On the Catasetum, with & the Researches of 206. Rosa canina, 110, 528, 530. Roser, 337, EE Sexual Forms of ial reference to win and others, LINN. JOURN.—BOTANY, VOL. XXVII. 554 INDEX. Rubiacesg of Fernando Noronha, 40. Ruscus, 279, 310. Saccharum officinarum, Linn., 72. Saco trapo, 482. Sagotia triflora, Duchass, & Walp., 29. Saleria, 13. Salices Synandre, 398. Salix, 333-457. acuminata, Sm., 339, 413-421. , Var. rugosa, Sm., 414. adscendens, Sm., 392. alba, Linn., 338, 343, 349, 354- 377. —— albescens, 374, 375. —., var, Russelliana, Sm., 3638, —, var. viridis, Wahlenb., 364, 365. alopecuroides, 353. alpestris, 4nd., 441. ambigua, Ehrh., 338, 392-394. —— —, var. genuina, 392. ——, var. longifolia, Wimm., 392. ——-, var. major, 392. 7x var. microphylla, Wimm., —, var. spathulata, Willd., 392, —— , var. undulata, 392, amygdalina, Linn., 347, 348. androgyna, 350. ——, var. porcellanea, Baenitz, 350. angustifolia, Wimm., 347, 359, 391, 417. angustissima, 449. aquatica, Borr., 381-289, 403. Arbuseula, Linn., 339-344, 396- 399, 410—440. —, erecta, Anders., 411. , var. foetida, Schleich., 411. ———, var. formosa, Willd., 411. ——, var. humilis, Anders., 411. ——, var. prunifolia, Anders., 411. —— var. thymelæoides, Schleich., 411. ——, var. vaccinifolia, 411. ——, var. venulosa, Anders., 411. "ei var. Waldsteiniana, Willd., arbutifolia, 433. arenaria, Linn., 496, 427, 498. argentea, Wimm., 370, 389, 427. atrocinerea, Brot., 280. aurita, Linn., 338, 343, 344, 377- 394, 417-452. Ve teen Wimm., 339, Morte “nigricans, Heidenr, 409, Salix aurita, var. opaca, Anders., 429. ——--, var. subaurita, Anders., 429. , var. viminalis, Wimm., 414, 417. Aurora, Lestad., 392. bicolor, Ehrh., ftnote 403. bigemmis, 348. borealis, Fr., 400, 433. britannica, B. White, 338, 368. ezrulea, Sm., 370. Calodendron, Wimm., 414, 415, 420. campestris, Fr., 400. canariensis, Sm., 379. Caprea as host of Corticium, 139. Caprea, Linn., 388, 343. 344, 377- 394, 401-408, 417-429, 450. cinerea, Wimm., 331, 386, 387. dasyclados, Wimm., 420. —— nigricans, 408. repens, Lasch., 338, 394. viminalis, Wimm., 414. capreola, J. Kern., 338, 387-389. carinata, Sm., 410-412. cinerascens, 421, 450, 451. cinerea, Linn., 334, 338, 343, 344, 377-390, 420-430. , var. brevifolia, Anders., 380. ——, var. genuina, 381. ———, var. latifolia, Anders., 380. , var. laxiflora, 380. ———, var. limosa, Lestad., 939, 429, 430, , var. longifolia, Anders., 380. ——-, var. microphylla, 380. ———, var. nigricans, 407, 409. , var. oleifolia, Sm., 384. , var. purpurea, Wimm., 450. , var. repens, Wimm., 338, ——., var. spuria, Wimm., 380. , var. viminalis, Wimm., 414- concolor, Koch, 347, 348. conformis, Schleich., 409. contorta, Crowe, 347, 348. coriacea, 339, 409, 410, 433. cotinifolia, Sm., 407. crassijulis, 347. Croweana, Sm., 398, 399. cuspidata, Schultz, 338, 360. Daphneola, Sch 426, 427. daphnoides, 348. dasyclada, Wimm., 416, 420, 421. decipiens, Hoffm.. 338, 348-361. dichroa, Döl, 340, 452. : Dicksoniana, Forbes, 399, 412, 413. Dicksoniana, Sm., 339. C discolor, Koch, 347, 348. - INDEX. i 555 Kovatsii, 353. Salix Doniana, Sin, 340, 342, 452, , Salix Jacquiniana, 432, » Var, latifolia, 453. | 7——- var. leiocarpa, 453. —-, var. linearis, 453. "cn var. lingulata, 453. elesgnifolia, Lausch, 449. erecta, 411. eriantha, 447, 448. ferruginea, G. And., 339, 414- finmarchica, Willd., 392. firma, Forbes, 407. foetida, Schleich., 410, 411. Forbiana, Sm., 449, 450. formosa, Willd., 411. fragilior, Anders., 375. fragilis, Linn., 337-377. ag alba, Wimm., 364, 365, — —, var. decipiens, Kock, 349. ae pendula, Fr., 364, 365, "ach var. porcellanea, Baenitz, —, var. Russelliana, Sm., 363- 365. ^ ——-, Var. triandra, Wimmer, ftnote Friesiana, And., 391. furcata, 447. fusca, Wimm., 389, 390. usca, Linn., var. parvifolia, 391. genuina, 368, 448. glabra, 374, 375, 443. glabrata, Lundstr., 426. glabrescens, 421. glauca, Sm., 426, 428, 429, 441. glaucescens, 450. gracilescens, 353. gracilis, 447, 448. rahami, Borr., 339, 431-439, 444, ftnote 446. grandifolia, Ser., 378, 379, 450. grisophylla, 409. hastata, 399. He etschweileri, Heer, 399. Helix, Linn., 447, 448, 450. helvetica, Vill., 339, 426-429. herbacea, 334-344, 424-426, 437- 7. var. subpolaris, Anders., t. hexandra, Ehrh., 338, 361. hippophaifolia, Thuill, 355, 357, Hoffmanniana, Sm., 347, 348. | holosericea, Willd., 414, 418-420. | bumilis, 41]. | Intricata, Leefe, 413. | Læstadiana, Harim., 429, 430. Lambertiana, Sm., 447, 448. lanata, Linn., 339, 344, 491—490. lanceolata, Sm., 355, 356, 358. Lapponum, Linn., 389, $43, 344, 426, 427-431, 440, 44], 446. —, var. arenaria, Linn., 426. ——, var. herbacea, 441. —, var. pseudo-glauca, Syme, 426. , vat Seegen, Se latifolia, And., 417. latifolia, Forbes, 339, 347, 948, 359, 406, 453. Taurina, Sm., 339, 402-404. laxiflora, Borr., 403. leiocarpa, 453. linearis, 453. lingulata, 453. livida, 399. ludificans, B. White, 339, 402, 405. lutescens, 4. Kern., 338, 383, 384, 388 MacNabiana, MacGillior., 494. macrophylla, Hartig, 500. macrostigma, 449. margarita, B. White, 339, 441. marrubiifolia, Tausch, 426. Micheliana, Forbes, 418. microphylla, 347, 359. d mollissima, Ehrh., 355, 357, 358. monspeliensis, 369. montana, Forbes, 364, 365, 376. Moorei, Watson, 339, 438-440. multiformis, Dei, 355, 356, 368. myricoides, 404. Myrsinites, Linn., 339, 343, 344, 410, 432-438, 446. — Myrsinites-nigricans, Wimm., 433. myrsinitoides, Fr., 433. nigricans, Sm., 338, 343, 344, 394- 403, 433, 434, 438-444. ——, var. repens, Heidenr., 338, 394. —, var. Weigeliana, 401. nitida, Wimm., 408. nivalis, Hooker, 443. norvegica (/’r.), And., 441. oleifolis, 381, 382, 384. o , 429, L paluitris, Host, 914, 375. pedicellata, Desf., 378, 379. pendula, Ser., 364-367, 376. pentandra, Linn., 338, 343, 359- 365, A Li 4 phylicifolia, Linn., 338, 343, 344, 361, 395-413, 434, 437, 438. _—, var. borealis, Fr., 400. —, var. campestris, Fr., 400. 556 Salix phylicifolia, var. INDEX. nigricans, Wimm., 338, 397-401. , var. protea, 400. plicata, Fr., 392. polaris, 438. Pontederana, Schleich., 450, 452. procumbens, 432, 433. protea, 400. prunifolia, Sm., 410, 411, 413, 437, 440. pseudo-glauca, Syme, 426. puberula, Doli, 408. pubescens, Lundstr., 426. punctata, Hartm., 433. punctata, Wahl., 433. punctulato-scabra, 358. purpurea, Linn., 340, 348, 344, 898, 399, 447-453. purpureoides, 449. ramulosa, Borr., 447, 448. Reichardti, 4. Kern., 338, 386, 387. repens, Linn., 938, 343, 344, 389- 395, 427, 452. —— myrtilloides, Wimm., 392. reticulata, Linn., 334, 340-344, 423, 424, 443-447. —, var. nivalis, 443. ——, var. typica, 443. retusa, Linn., 438 rhætica, Kern., 399. rosmarinifolia, Koch, 390, 391. » var. angustifolia, Wulf., 891. rubra, Huds., 340, 342, 358, 398, 448-451. genuina, 449. rugosa, Leefe, 414, 417-420. Russelliana, Sm., 363, 367-376. Sadleri, Syme, 339, 422, 423. saxetana, B. White, 339, 434, 436. Schraderiana, Willd., 395. sejuncta, 444. somireticulata, B. White, 340, 444, 46. sericans, Tausch, 339, 414, 415, 417. sericea, 443, 447-450. serrata, 433. serta, B. White, 339, 436. sibyllina, B. White, 340, 446. silesiaca, Wimm., 378, 379. simulatrix, B. White, 339. Smithiana, Willd., 339, 358, 413- ——— Var. pseudo-stipularis, 416. sobrina, B. White, 339, 440. soluta, 444. sordida, Kern., 340, 0-452, Salix sordida, var. rubella, 451. spathulata, Willd., 392. speciosa, Host, 353. sphacelata, Sm., 386. splendens, Bray, 310. spuria, Schleich., 339, 430, 431. Stephania, B. White, 339, 424, 426. stipularis, Leefe, 413-418. stipularis, Sm., 339, 413-418. strepida, Schleich., 339, 408-410. , var. nitida, Wimm., 408. —-—, var. puberula, Doll, 408. , var. vaudensis, Forbes, 408. Stuartiana, Sm., 426. styligera, 447, 448. subaurita, 429. subcinerea, And., 430. ` subdola, B. White, 338, 354, 355. submyrsinites, 433. subnigricans, 433. subobscura, And., 417,418, 419. subpolaris, Anders., 437. subpurpurea-cinerea, Kern., 450. subtriandra, 353. superata, B. White, 339, 423. tenuifolia, Sm., 402. tenuijulis, 347. tenuior, Borr., 402. tephrocarpa, Wimm., 339, 402, 406, 410. tetrapla, Walk., 398, 401. . Trevirani, Spreng., 347, 348, 355- 359 triandra, Linn., 338-361. , var. alba, 355. , , Var. amygdalina, Linn., 347. , var. angustifolia, 347. , Var. concolor, Koch, 347. , var. contorta, Crowe, 347. —, var. crassijulis, 347. —— , var. discolor, Koch, 347. —— , var. Hoffmanniana, Sm., 347. , var. tenuijulis, 347. , var. "I revirani, Spreng., 347. ———, var. Villarsiana, 347. ,var, viminalis, Wimm., 355, 956, 358. ——, var. vulgaris, 347. typica, 443. undulata, Ehrh., 338, 349, 353- 358. vaccinifolia, Walk., 411, 440. vaudensis, 408. velutina, Schrad., 339, 414-420. venulosa, Sm., 410, 411. vestita, 374. Villarsiana, 347. INDEX. 557 Salix viminalis, Linn., 339, 343,344, 348, 355, 391, 413-421, 448—451. —, var. Caprea, 416. —-—, var. intricata, Leefe, 413. ——-, var. purpurea, 448. ~, var. repens, Losch, 391. Th var. stipularis, Leefe, 413, viminaloides, 449. viridis, Fr., 338, 363-377. » Var. albescens, 374. ; var. excelsior, 374, 375. ———, var. glabra, Wimm., 374. — —, Var. vestita, Wimm., 874. ——-, var. viridis, Wimm., 374. vitellina, Linn., 338, 349, 370, vulgaris, 347, 390. Wahlenbergii, And., 339, 423. Waldsteiniana, Wilid., 411. Wardiana, B. White, 339, 402- 405. Weigeliana, Willd., 400. Woolgariana, Borr., 447, 448. Santalacerm of Patagonia, 498. Sapindaces of Fernando Noronha, 25. Sapium, 4, 12. biglandulosum, 61. sceleratum, Ridley, 5, 60, 95. Sapotacese of Fernando Noronha, 43. arcostemma Gilliesii, Decne., 494. tncanum, Decne., 494. Sargassum vulgare, 4g., 78. Saxc-Gothgea, 235, 249, 293-328. Saxifragere of Patagonia, 483. Schizolepis Braunii, 315. Schizophycee, 463. S pirophyllum commune, Fr., 80. Schmidelia, 12. insulana, Ridley, 5, 25. Schultesia stenophylla, Mart., 45. Sciadium, 4. Braun, 462. Sciadopitys, 227, 935, 240, 241, 258, | 258, 276-281, 296-326. Scirpus micra»tbus, Vahl, 67. silvaticus, 136. paria dulcis, Linn., 3, 13, 51. purpurea, Ridley, 3, 51. Scrophuluriaces: of Fernando Noronha, 5; of Patagonia, 496. Scutellaria numniulariæfolia, Hook. f., Sedum Telephium, 531. Selaginella, 247, 298. Senecio Hualtata, Bertero, 490. linarieefolius, Poepp., 490. mendocinus, Phil., 490. miser, Hook. f., 489. ranconensie, Sch.-Bip., 490. subulatus, D. Don, 490. Sequoia, 235, 237, 250, 296-328. gigantea, 231, 287, 241, 243, 256, 257, 320. sempervirens, 243, 304. Sesbania ægyptiaca, Pers., 28. Sesuvium, 10. distylum, Ridley, 3, 38. portulacastrum, Linn., 38. Setaria caudata, Roem. d Schult., 71. scandens, Schrad., 9, 71. Sida altheæfolia, Sw., 5, 23. atrosanguinea, Jacq., 23. carpinifolia, DO., 23. glomerata, Cav., 23. paniculata, Linn., 28. spinosa, Linn., 28. sulphurea, 4. Gray, 478. Siphonocladiaceæ, 526. Sitoplidus oryzæ, 16. Solanaceæ of Fernando Noronha, 48; of Patagonia, 495. Solanum, 11, 50. botryophorum, Ridley, 50. corniculatum, 50. cornigerum, André, 50. Dulcamara, 51. eleagnifolium, Cav., 496. mammosum, var. corniculum, André, 50. nigrum, Var., Sendtn., 49, 50. oleraceum, Dunal, 11, 49. iculatum, Linn., 5, 50. Neaforthiæ, André, 50, 51. Sorghum balepense, Pers., 72. vulgare, var. saccharatum, 72. Spatoglossum, 466, 467, 468. Spermacoce parviflora, Hemsl., 40. Sphacelaria, 469. Sphæralcea, 478. bonariensis, Griseb., 478. cisplatina, St.-Hil., 478. collina, Phil., 478. coquimbana, Phil., 478. flexuosa, Gill, 478. obtusiloba, G. Don, 478. Spheria, 111. Spheroplea, 458. annulina, 4g. 458. Spigelia anthelmia, Linn., 44. Spirogyra, 408, 528. —— | Spondias purpurea, Linn., 5, 11, 27. Spondylantha aphylla, Presl, 27, Statice brasiliensis, Boiss., 493, 494. californica, 494. caroliniana, Wait., 494. chilensis, Linn., 494. Limonium, Linn., 494. | Steira satureifolia, Sch.-Bip., var. an- gustifolia, Baker, 486. Stellatostroma, 110. er ei tae 558 INDEX. Sterculia fetida, Linz., 24. Sterculiaces of Fernando Noronha, 24. Stereum, 95, 157-164, 182-187, 196, abietinum, Fr., 115. acerinum, Fr., 202. acerinum, Berk. & Broome, 130. adustum, Lév., 195. - affine, Lév., 172. albo-badium, Fr., 194. albo cinetum, Berk. § Broome, alliciens, Berk. § Cooke, 201. alutaceum, Berk. § Cooke, 162. amenum, Massee, 193. annosum, Berk. & Broome, 202. anemum, Kalchb., 182. Aratze, Speg., 201. arenicolum, Berk., 201. | aterrimum, Cooke, 183. | atro-zonatum, Speg., 177. | attenuatum, Lév., od. auriusculum, Berk. & Broome, avellanum, Fr., 103. | badio-ferrugineum, Mont., 101, balsameum, Peck., 196. bellum, Massee, 177. | Berkeleyanum, Mont., 101. bicolor, Fr., 177, 183, 190. bizonatum, Berk. § Curt., 178. Bolleanum, Mont., 166. Boltonii, Sace., 115. Boryanum, Fr., 175. ! Cacao, Berk., 100. | calyculus, Berk. d Curt., 166. | candidum, Fr., 200. candidum, Schwein., 202. caperatum, Massee, 161. carolinense, Cooke & Rav., 167. chartaceum, W. Mey., 196. cinerascens, Massee, 179. cinereum, Lév., 189. citrinum, Berk. & Curt., 203. Coffearum, Berk. & Curt., 194. coffeatum, Berk. & Curt., 190. complicatum, Fr., 178, 181, 188. concolor, Jungh., 178. conspurcatum, Massee, 196. contrarium, Berk., 188. corrugatum, Massee, 204. crenatum, Lév., 169. cristatum, Berk. § Curt., 167. crocatum, Fr., 105. crucibuliforme, Massee, 168. Curtisii, Berk., 195. curtum, Fr., 166, 168. | cyathiforme, Fr., 159, 205. | eyathiforme, Currey, 168. | cyphelloides, Berk. 4 Curt., 172. | Stereum damecorne, Link, 96. desolat ionis, . 179. diaphanum, Cooke, 162. disciforme, Fr., 189. effusum, Berk., 179. elegans, Fr., 161, 162, 163, 165, 174. elevatum, Berk. d: Cooke, 100, 205. endocrocinum, Berk., 179. endoleucum, Berk. & Broome, 202. fasciatum, Fr., 180. ferreum, Berk. § Curt., 197. ferrugineum, Fr., 103. fissum, Berk., 169. frustulosum, Fr., 199. frustulosum, Karst., 204. fuliginosum, Pers., 199. fulvo-nitens, Berk., 162. fulvum, Massee, 180. Galeottii, Berk., 176. gausapatum, Fr., 180. glabrescens, Berk. §- Curt., 169. glabrum, Massee, 177. glaucescens, Fr., 199. Goliath, Speg., 159, 160. Haydeni, Berk., 199. hirsutum, Fr., 181-201. , var. cristulatum, Quelet, 181. —., var. subcostatum, P. Karst., 181. hydrophorum, Berk., 159, 160. hylocrater, Speg., 160. illudens, Berk., 181. induratum, Berk., 196. insulare, Berk. 4. Broome, 200. involutum, Klotzsch, 176. Kalchbrenneri, Massee, 182. letum, Berk. & Curt., 185. levigatum, Speg., 200. , var. mesopoda, Speg., 200. Leichardtiana, Massee, 175. Lepra, Berk. & Broome, 130. leucopheum, Lév., 182. ` Leveillianum, Berk. d Curt., 197. lobatum, Fr., 170, 175, 176, 188, 190. lugubris, Cooke, 182. luteo-badium, Kee, 103. luteo-badium, Fr., 175. medicum, Curr., 202. Mellisii, Berk., 163. membranaceum, Fr., 177. 8 Micheneri, Berk. 4 Curt, 172, 183. 199 micraspis, A . Mique Be SP font. 161. molle, Massee, 175. Moselei, Berk., 166. INDEX. 559 Stereum multizonatum, Berk. / Broome, | Stereum rugosum, Fr., 123, 126, 171 167. | 189-192. iu mytilinum, Fr.. 170. Nicarague, Berk. & Curt., 183. nigricans, Lév., 173, 183. nitidulum, Berk., 161, 163, 164. notatum, Berk. § Broome, 197. obliquum, Mont. § Berk., 170. ochraceo-flavum, Massee, 184. ochraceo-flavum, Fr., 192. ochroleucum, Fr., 184. odoratum, Fr., 198. Ostrea, Nees, 175, 176, 178, 188. ostreatum, 197. pannosum, Cooke, 185. paraguariense, Speg., 195. partitum, Berk. & Broome, 108. percome, Berk. & Broome, 185. pergamenum, Berk. & Curt., 161. Pertatum, Berk., 175. petalodes, Berk., 165. phalenarum, Kalchbr., 182. pheum, Berk., 98. pictum, Berk., 185. princeps, Berk., 171. princeps, Jungh., 170, 171, 183. prolificans, Berk., 167. pruinatum, Berk. / Curt., 198. pulchrum, Cooke, 184. pulchrum, Schw., 98, 99. pulverulentum, Massee, 174. pur q roum, Pers., 120, 182, 186, ——, var. lilacinum, Fr., 186. ———, var. venosum, Quelet, 186. pusillum, Berk., 166, 174. pusioluin, Berk, § Curt., 168, 169. quercinum, Berk., 182. radians, Fr., 188, 205, radiato-fissum, Berk. § Broome, radiatum, Peck, 195. radicale, Massee, 187. rameale, Massee, 187. Ravenelii, Berk. § Curt., 164, 165, 205. reniforme, Fr., 96. retirugum, Cooke, 186. rheicolor, Mont., 98. rhicnopilus, Massee, 188. igen P. Karst., 189. rigidulum, 4 174. rigidum, LP as rimosum, Berk., 187. Rivulorum, Berk. 4: Curt., 167. ruberrimum, Berk. g Broome, 198. | | | | | | ET var. aurantiacum, P. Karst., sanguinolentum, Fr., 189, 191. Sarmienti, Speg., 190. Schomburgkit, Berk., 115. Schraderi, Thuem., 174. Schulzeri, Quel., 203. scriblitum, Berk. & Cooke, 174. scytale, Berk., 171, 183. seriatum, Berk. A Curt., 194. sericeo-nitens, Speg., 191. sericeum, 188. simulans, Berk. g Broome, 189. siparium, Massee, 204. Sowerbeii, Massee, 164. spadiceum, Fr., 182, 190. sparsum, Berk., 203. spathulatum, Berk., 171. spectabile, K/otsch, 172, 193. Spegazzianum, Massee, 192. spongiosum, Massee, 172. jprucei, Berk., 175. stratosum, Berk. § Broome, 208. striatum, Fr., 174. strumosum, Fr., 203. subcostatum, P. Karst., 181. subcruentatum, Berk. d' Curt., 171. subpileatum, Berk., 171, 192, 197. eubpurpurascens, Berk, § Broome, 1 sulfureum, Fr., 203. sulphuratum, Berk. & Rav., 192. surinamense, Zév., 161. tabacinum, Fr., 112. tenerrimum, Berk. § Rav., 165. tenuissimum, Berk., 102. Thoretii, Berk., 165. . triste, Berk. & Curt., 192. Tuba, Berk. g- Broome, 165, 205. tuberculosum, Fr., 204. tumulosum, P. Karst., 204. umbrinum, Berk. & Curt., 104, 113, 193. variolosum, ., 198. vellereum, mad 173. versicolor, Fr., 172, 177, 185. versiforme, Berk. d Curt., 193. vespilloneum. Berk., 173. vibrans, Berk. & Ourt., 117. villosum, Zév., 173. vinosum, Quelet, 186. vitile, Fr., 193. vorticosum, ZY., 194. xanthellum, Cooke, 164. Zippelius, Jungh.. 173. rufum. Py. 1987 anes Stifftia since, ee. E 491. i t, | Stipitate enoc , 96. , "Së erch Berk. 4 Cur 3 i 749. 560 Strobus, 281. Strombocarpa, 483. Surirellez of Fernando Noronha, 84. Swartzea pennata, 19. Swartzia pennata, Webster, 34. pinnata, Willd., 19, 34. Synandrex. 337, 340-347. Synedra Acus, Kuetz., 83. affinis, var. delicatula, Grun., —, var. hybrida, Grun., 83. lanceolata, Kuetz., 83. nitzschioides, Grun., 83. oxyrhynchus, Kwetz., 83. tabulata, Kuetz., 83. Syringa vulgaris, 122. Tabellarie of Fernando Noronha, 84. Tachytes inconspicuus, 13. * Tajuja,” 36. Talinum patens, Willd., 10, 22. “ Tamiarana,” 60. Taonia, 464, 466. Taxaces, 309, 322. Taxes, 235. Taxodiex, 235, 303. Taxodium, 235, 250, ftnote 255, 322. distichum, 231, 256. ———, var. imbricaria, 256. sempervirens, 284. Taxus, 235, 249, 296-328. baccata, ftnote 255, 257, 295, tardiva, 311. emnoceras vesiculosus, 13. Temperatæ, 337. Tephrosia cinerea, var. littoralis, enth., 28. Terminalia Oatappa, Linn., 27. Terminaliopsis, 1, 2, 27, 28. Thelephora, 95, 117, 153-160. ietina, Pers., 115. acerina, Pers., 202. adusta, Lév., 195. albo-badia, Schw., 194. albo-carnea, Schwein., 142. alutacea, Schrad., 139. anthochroa, Pers., 141. attenuata, Lév., 103. aurantiaca, Berk. & Mont., 162, badia, Hook., 170. ia, Kunze, 100. ella, Kunze, 177. bicolor, Pers., 178. bombycina, Berk., 136. cerulea, Berk., 151. candida, Schw., 200. cariosa, Pers., 132. INDEX. . Thelephora caulina, Schw., 141. — Cerasi, Pers., 105. cinerascens, Schw., 179. comedens, Nees, 159. corrugata, Fr., 110. corrugata, Lév., 204. corticalis, Schrad., 124. cruenta, Alb. & Schw., 112, 118, 122. cyathiformis, Fr.. 159. mecorne, Fr., 96. decorticans, Pers., 155. diaphana, Schwein., 162. elegans, Fr., 162. epidermea, Pers., 133. epispheria, Schwein., 111. erinacea, Jungh., 173. fasciata, Schw., 180. fastidiosa, Berk. & Broome, 131. floriformis, Schwein., 162. frustulosa, Fr., 199. fulva, Lév., 180. fusco-purpurea, Pers., 109. gausapta, Fr., 180. labra, Lév., 177. lvelloides, Schwein., 154. hirsutum, Willd., 181. imbricatula, Schw., 103. incrustans, Pers., 127. juniperina, Fr., 124. juratensis, Pers., 103. lactea, Fr., 132. leta, 99. levigata, Schw., 107. Laurocerasi, Berk., 191. laxa, 155. Leichardtiana, Lév., 175. lilacea, Rabenh., 156. lilacina, Pers., 186. livida, Sommerf., 151. Lycti, Pers., 122. molle, Fr., 143. molle, Lév., 175. Mougeotii, Fr., 112. — nigrescens, Schrad., 155. nigricans, Lév., 183. ochraceo-flava, Schw., 184. odorata, Fr., 198. Padi, Pers., 110. pallescens, Schwein., 130. populina, Sommerf., 121. pulchra, Schwein., 185. pulverulenta, Lév., 175. & purpurea, Cooke & Morg., 115. purpurea, Hussey, 186. puteana, 144. rheicolor, Montag., 99. rhicnopilus, Lév., 188. rubiginosa, Schrader, 97. rubropallens, Schweiu., 145. INDEX, Thelephora salicina, Pers., 138, sanguinolenta, Alb. E Schw., 189. sarcoides, Fr., 122. sebacea, Fr., 127, sera, Pers, 127. seriata, Fr., 126, Sowerbeii, Berk., 164, 167. spadicea, Fr., 190. speciosa, Fries, 97. spongia, 204. tabacina, Fr., 112. Ulmi, Lasch., 157. versicolor, Swartz, 173. versiforme, Fr., 126. Th itieola, Schwein., 146. eiephorez, a Monograph of the, b G. Massee, 95. grap d Thuiscarpus juniperinus, 311. Thuja occidentalis, 109. Thuya, 229, 935, 254, 257, 981, 286, 288, 298, 302, 322-324. Bodmeri, 255. filifera, 256. gigantea, 231, 237, 240, 243, 257, 266, 287-289. Mertensiana, 231. obtusa, 256, 302. occidentalis, 288, 289. orientalis, 281, 288. pisifera, 229, 264. licata, 256. areana, 256, Thuyopsis, 250, 322. borealis, 289, 309. dolabrata, 289, 302. Thysanodactylus lineatus, 19. Tiliaceg of Patagonia, 479. Tilopteridese, 463-470. Tilopteris globosa, 469. Torreya, 235, 249, 293-300. grandis, 323. Tortula, sp., 74. Tragia volubilis, Linn., 60. Trametes hydnoides, 125, 183. Triandre, 337, 338, 355, 358. nanosperma racemosa, Griseb., 37. . . Tajuja, Mart., 36. Triceratium alternans, Kären, 85. elegans, Grev., 85. favus, Ehrend., 85. Hardmanianum, Witt., 85. pentacrinus, 7. Wallich, 85. „ . „trisulcum, Bailey, 85. Trichocline incana, Cass., 492. Tripsacum hermaphroditum, Linn., 12. Tropicæ, 337, Tsuga, 236, 309. Brunoniana, 249, 315. canadensis, 237, 247, 249, 325. Hookeriana, 249. LINN. JOURN.—BOTANY, VOL. XXVII. [ Tsuga Mertensiana, 257. Sieboldi, 249. Tussilago Farfara, 528. Typha latifolia, 137. Ulva, 465. lactuca, Linn., TY. lobata, Kuetz., 79. Ulvaces, 526. Urasu, 481. . Urena lobata, Linn., 12, 23. Urticacez of Fernando Noronha, 62. Ustulina vulgaris, Zul., 81. Utricularia, 13. Vacciniaceæ of Patagonia, 493. Vaccinium, 120. Valonia filiformis, Dickie, 78. ventricosa, J. 4g., 78. ** Vassorinha," 51. Vaucheria, 524. Vegetable Biology, Studies in.—VI. An Investigation into the True Nature of Callus: The Vegetable-Marrow and Ballia callitricha, Ag., by Spencer le M. Moore, 501-526. ——. VII. Some Michrochemical Reactions of Tannin, with Re- marks upon the Function of that Body and its Excretion from the General Surface of Plants, by Spencer le M. Moore, 527-538. . : Vegetable Marrow and Ballia calli- tricka, Ag., Spencer Moore, 501-526. Velellas, 10, 18. Verbena, 497. Berterii, 497. bonariensis, Linn., 497. bryoides, Phil., 498. pulchella, Sweet, 497. tenera, Spreng., 497. thymoides, Phil., 497. Verbenaceæ of Fernando Noronha, 52; of Patagonia, 497. Verdu larga, 477. Vesicaria andicola, Ġill., 476. arctica, Hook., 476. mendocina, Phil., 476. montevidensis, Eichl., 476. Vicia graminea, Linn., 481. littoralis, Jacq., 28. ` Viminales (Salices), 335, 337, 339, 355, 358, 413. Vinea rosea, Linn., 45. Vireo, 4. Virescentes (Salices), 337. Vitalba, sect. of Clematis, 475. Vitis, 11, 142, 524. sicyoides, Baker, 26. 28 56] 562 Vitis vinifera, Linn., 26, 502. Vulticeps, 116. Berkeleyi, Cooke, 117. Waltheria americana, Linn., 24. Welwitschia, 251, 285. White, Dr. F. Buchanan, A Revision of the British Willows, 333-457. Widdingtonia, 250. Wilbrandia villosa, Cogn., 485. Willqws, A Revision of the British, by Dr. F. Buchanan White, 333-457. Wissadula hirsuta, Presl, 23. Xylaria polymorpha, Grev., 81. Xerocarpus levissimum, P. Karst., 132. INDEX. Zamia, 296, 297. Zanardinia marginata, J. 4g., 76. Zanonia discolor, 64. Zea Mays, Linn., 71. Zenaida Noronhæ, 4, 12. Zephyranthes, 473, 500. Andersoni, Herb., 499. filifolia, Herb., 500. n. sp., Ball, 500. Zonaria, 468. lobata, 4g., 77. Zornia, 9. diphylla, Pers., 29. —, var. elatior, Benth., 29. —, var. reticulata glabra, Benth., 29. Zygnema, 523. . Zygophylles of Patagonia, 479. END OF THE TWENTY-SEVENTH VOLUME. PRINTED BY TAYLOR AND FRANCIS, RED LION COURT, FLEET STREET. SL Notes on th ac i Botany of Fernando No ha. Rrpzxy, M.A., RS du um Byn ` MISSOURI BOTANICAL GARDEN. SE WILLIAM WESLEY & Sen, Booksellers & Publishers, Te Street, Strand, “LINNEAN SOCIETY OF LONDON. LIST OF THE OFFICERS AND COUNCIL. - Elected 24th May, 1889. DH PRESIDENT. William Carruthers, F.R.S., F.G.S. : VICE-PRESIDENTS. ohn. irem Suh FBS. _ Robert Braithwaite, M.D. i Frank Orisp, LL.B., B.A. TREASURER. Frank Crisp, LL.B., B.A. SECRETARIES. | W. Percy Sladen, F.G.S COUNCIL. Prof. P. Martin Duncan, M.B., F.R.S.. Prof. G. B. Howes, FZS. Russell, M.P. D. H. Scott, M.A., Ph.D. W. Percy Sladen, F.G.S. :John Jenner Weir, RS, LIBRARIAN AND ASSISTANT SECRETARY. ^ James Edmund Harting, F.Z.S. "LIBRARY COMMITTEE. j This consiste of nine Fellows (three of whom retire annually) and of the four Officers ex officio, in all thirteen members, The former are elected annually: -the SN in June, and serve till the succeeding Anniversary. The ‘meet at 4 P.u., at intervals Saring the Session. The Sr for ,in gege to the officers, . Ernest Clarke, es | William Sweetland Dallas, Ber TR G. B. Howes, RSR ` F. W. Oliver, ac iess FES. Price 6s. THE JOURNAL. THE LINNEAN SOCIETY. - CONTENTS. | "us I. Life-History of a Stipitate Freshwater Alga ere Groraz Masser. (Commanicated by the Beete ue (Plate XIL) .. . 457 . I. On the Systematic Position of the Dictyotaces, with. E special reference to the Genus Dictyopteris, Lamour. ~~ By Tuomas Jonnson, B.Sc. (Lond.), University — Scholar in Botany, Demonstrator of Botany in the ~ Normal School of Science. (Communicated by D. H. Scorr, M.A., Ph.D., F.L.S., Assistant Professor of my in the Normal School of AIMO. (Piate III. Further Contributions to the Flora of onia. - Bro ien the late Jony Barr, F.R.S., M.R.L.A., F. e... an V. Studies in Vegetable Biology —VI. An Seite: i into the True. Nature of Callas :—The Vegetable Marrow and Ballia callitricha, Ag. By SPENCER E Le M. Moos, F.L.S. (Plate XIV.) .. 501 . V. Studies in Vegetable Biology. — VI. ‘Some Miero- chemical Reactions of Tannin, with Remarks upon the Function of that Body and its Excretion from the tsch Surface of Plants. By Geer Mess, Lag Index, Titlepage, Contents, ‘be, to Vol. XXVII. See Notice on last page of Wrapper. LONDON: SOLD AT THE SOCIETY'S APARTMENTS, BURLINGTON HOUSE, Ee PICOADILEY, Wy o m ot LINNEAN SOCIETY OF LONDON. LIST OF THE OFFICERS AND COUNCIL. .. ' Elected 24th May, 1890. PRESIDENT. Professor Charles Stewart, M.R.O.S. VICE-PRESIDENTS. J. Q. Baker, F.R.S. William Carruthers, F.R.S. Frank Crisp, LL.B., B.A. Prof. P. Martin Duncan, F.R.S. TREASURER. Frank Crisp, LL.B., B.A. SECRETARIES. B. Daydon Jackson, Esq. | W. Percy Sladen, F.G.S. COUNCIL. ‘John Gilbert Baker, F.R.S. B: Daydon Jackson, Esq. ... Alfred William Bennett, M.A., B.Sc. | John W.S. Meiklejohn, M.D. =- Robert Braithwaite, M.D., M.R.O.S. | George R. M. Murray, Esq. ^F. Herbert Carpenter, D.Sc., F.R.S. | Edward B. Poulton, M.A., F.R.S. + Wiliam Carruthers, F.R.S., F.G.8. Dukinfield H. Scott, M.A., Ph.D. Frank Crisp, LL.B., B.A. David Sharp, M.B., F.Z.8., F.ES.- P. Martin Duncan, F.R.S., F.G.8. W. Percy Sladen, F.G.S. Charles Stewart, M.R.C.S. LIBRARIAN AND ASSISTANT SECRETARY. James Edmund Harting, F.Z.S. LIBRARY COMMITTEE. This consists of nine Fellows (three of whom retire annually) and of the four officers ex officio, in all thirteen members. The former are elected annually by the Council in June, and serve till the succeeding Annive . The Committee meet at 4 r.m., at intervals during the Session. The Members for 1890-91, in addition to the officers, are :— Alfred William Bennett, M.A., B.Sc. | St. George Mivart, M.D., F.R.S. George 8. er, F.G.S. Francis Wall Oliver, B.A. -Ernest Clarke, Esq. Francis P. Pascoe, Esq. _ . Prof. George Bond Howes, F.Z.S. Henry Seebohm, F.Z.8. es Prof. H. Marshall ard, MA, F.R.S. orx,—The Charter and Laws of the Society, a3 amended to E the 19th March, 1891, may be had on application. May 8. Vor. XXVII. "BOTANY. MISSÓURI. BOTA NICAL GARDEN. CONTENTS. WILLIAM WESLEY & Sox, Booksellers & Publishers, 28, Essex Street, Strand, -LINNEAN SOCIETY OF LONDON. ' LIST OF THE OFFICERS AND COUNCIL. Elected 24th May, 1889. PRESIDENT. ` William Catruthére, F.R.S., BGS. vh CE- PRESI DENTS. chn Anderson, MD., F. BS, Robert Braithwaite, M.D aeren FRS. Wah "Frank Crisp, LL.B., B.A. . TREASURER. ‘Frank Crisp, LL.B., B.A. SECRETARIES. | W. Percy Sladen, F.G.S. COUNCIL. Prof. P. Martin Duncan, M.B., F. RS. Prof. G. B. Howes, F.Z.S. Game Ee e Esq. . Murray, Or DH Soot MAS PhD. D. 8 bo? MA, FRS |W. Percy Sladen, F.G.S. E ohn Jenner Weir, F.Z.S. LIBRARIAN AND ASSISTANT SECRET ARY. James Edmund Harting, F.Z.8. LIBRARY COMMITTEE. This consists of nine Fellows (three of whom retire annually) and of the four licers ex officio, in all thirteen members. The former are elected annually the Council: in June, and serve till the succeeding Annive The omr meet at 4 pr, at intervals during the Session. The Mem ers for x 1889-90, in addition to the officers, are :— Alfred W. Bennett, M.A. . | Ernest Clarke, A : Boulger, F .&.8. -< | William Sweetland Dallas, Esq. Esq. Prof. G. B. Howes, RSR, rtr Gia, Dës, F.R.S. | F. W: Oliver, B.A., Dës, "Francis P. Pascoe, PRS. - LINNEAN SOCIETY OF LONDON. LIST OF THE OFFICERS AND COUNCIL. Elected 24th May, 1890. PRESIDENT. Professor Charles Stewart, M.R.O.S. VICE-PRESIDENTS. ¥ '&. Baker, FRS. ` : William Carruthers, F.R.S. Frank 'LL.B., B.A. Prof. P. Martin Duncan, F.R.S. TREASURER. Frank Crisp, LL.B., B.A. SECRETARIES. | W. Percy Sladen, F.G.S. COUNCIL. B. Daydon Jackson, Esq. ‘John W.B. Meiklejohn, M.D. R. M. Murray, Ke W Pony WE _ Charles enen 1 MEOS ` LIBRARIAN AND ASSISTANT SECRETARY. ` dames Edmund Harting, F.Z.8, A » LIBRARY OOMMITTEE. Sage ZE nine Fellows (three of whom retire annually) and of the four ^ officers ae t. ar eficio in all thirteen members. 'The former are elected annually thee Council in June, and serve till the succeeding Anni The .meet at 4 P.u., at intervals during the Session. The Mem ers for ls in addition to the officers, are :— Mivart, E F.RS.. LINNEAN SOCIETY OF LONDON. LIST OF THE OFFICERS AND COUNCIL. Elected 24th May, 1890. d PRESIDENT, CH Professor Charles Stewart, M.R.O.S. Axe VICE-PRESIDENTS. d G. Baker, F.R.S. — William Carruthers, F.R.S. Frank Orisp, LL.B., B.A. Prof, P. Martin Duncan, F.B-8. 3 TREASURER. Frank Crisp, LL.B., B.A. whe SECRETARIES. .,. B. Daydon Jackson, Esq. | W. Percy Sladen, F.G.8. ' COUNCIL. Lon ve? Jackson, Esq. 8. Meiklejohn, MD. R. M. M , Esq. Herbert r, D. À Edward B. Poulton, , E.R.S. illian Oarru a, F.R.8., F.G.8. Dukinfield H. Scott, M.A., Ph.D. doris Ze B.A. David , M.B., RSR, F.E.S. en J n, F.R.S., F.G.8. W. Percy iden, F.G.8. ETE Charles Stewart, M.E.C.8. LIBRARIAN AND ASSISTANT SECRETARY. James Edmund Harting, F.Z.8. LIBRARY COMMITTEE. : * ` This consists of Nei Fellows (three of whom retire annually) and of the four officers ex officio, in all thirteen members. The former are elected annually oa the Council in June, and serve till the succeeding Annive versary. Tao E af d rite at intervale during the Session. The Members for ` 1890-91, in addition to the officers, are:— E ‘Alfred William Bennett, M.A., BSc. | St. Mivart, M.D., F.R.8. OU B. , ROB Se = | Fran Pus all Oliver, B.A. ATG Howe PZS dr Zeg l rae ee of the ct my eli amended A RULES FOR BORROWING BOOKS zap THE LIBRARY. ge As amended by the Council, 15th March, 1888. 1. No more than Six volumes shall be lent to one person at the same time without the spoin bare of the un or one of the Secretaries. | P^ 2. All books shall be returned before the ajia ot | Six weeks from the time of their being taken out, but ‘if not required by any other Fellow, they may, on spplication, » š kept for a further period of Six weeks. fe 3. All books lent shall be regularly atest: w ae Librarian in a book appropriated for that purpoeec P E 4. No work forming part of Linneus’s own rary shi be lent out of the Library under any circ f Norz.— Certain other works ore included în this prohibition, a such as costly illustrated works, and volumes belonging to Ee sets which could not be replaced $f lost. ot The GENERAL INDEX to the first twenty volumes of the Journal (Boraxx) and the Botanical portions of the Proceedings. from November 1888 to June 1890 is now ready for delivery to | the Fellows on Application. It may be had in cloth, or in sheets for binding. Fallows irè y desired, when ME I 6 NOTICE. Se XXVI, XXVIL, and XXVIII. are in course of con- ent emer to vespere the publication of papers, and the „desire to x THE LIBRARY. — As amended by the Council, 15th anc: T ; 1. No more than Six volumes shall be lent a or one of the Secretaries, — — ae Ste A All books shall be returned before the required by any other Fellow, they may, on pln, be kept for a farther period of Six Notit 5 : Librarien i in a book appropriated for that purpose. ` A No work forming part of Linneus’s own Libr - be lent out of the. : rary under Norz.— Certain other. works are eluded in this such as costly illustrated works, and volumes gts. sets which could not be Pg if lost. The GENERAL INDEX to the first twenty volumes of the . Journal (Boraxy) and the Botanical portions of the Proceedings S from November 1838 to June 1886 is now lae for nmre d the Fellows on Application. ` od ` It may be had in deti a: dei for binding. ‘Fellows are “a : desired, when SE for See © a state in which form. hi | 04 NOTICE. | E Vols. XXVI. and XXVII. are in course of concurrent issue SEN the Ver of papers, and the- Parts already RULES FOR BORROWING BOOKS FEO THE LIBRARY, : 25 Ae amended by the One 21508 uen 1898. E. : 1. No more T Six volumes shall be lent to ‘one | je rs at the same time without z a d = ee 2 required by any other Fellow, ry ads may, on inn be ‘kept for a further period d * Be CMM. de ar vum * Librarian i ina book propriate f ii that E Nomz.— Certain other nis are included in this pr such as costly illustrated works, and volumes sete which could not be replaced if lost. ` The GENERAL INDEX to the first eet E of the SH Journal (Boraxx) and the Botanical portions of the pens from November 1888 to June 1886 i is ae verd for delivery ec the Fellows on Apri ue ; Së Ns x NOTICE. SS | i p | ; Vols XXVE and XXVIL are in course of concurrent issue, already published being as under :— Von, XXVI. Nos..173-174. br (Nos. 175-180 inclusive are reserved for the completion. of Messrs. Forbes and Hemsley’s “Index Flore ` . Sinensis. ”) OL. xxvi. A 181, and the eurrent number, 182. . 1 to this ani is specially requested, to prevent application for unpublished numbers.. b SCH Meetings of the present Session will be held ag ` 1890, Saturday ^| 1890. Thursday, June 5 Së, 24 | 19 3? 3? The Chair will be taken at Exaur o'clock in the Evening pre- ely at every Meeting, excepting on the 24th May, the day nted for the Anniversary Elections, when the Chair will be at Tanzs o'clock i in the afternoon. RULES FOR BORROWING. BOOKS FRON THE LIBRARY. E As amended by the Councit, 155 March, x ix | l. No more than Six volue shall be lent to” ‘one’ oe SC St the same time without the special leave of the Council E or one of the Secretaries, — . 2. All books shall be returned before the. expiration of ` Six weeks from the time of their being taken out, but if not required by any other Fellow, they may, on application, be kept for a further period of Six weeks. 3. All books lent shall be regularly entered by the - Librarian in a book appropriated for that purpose. 4. No work forming part of Linnsus's own Library shall ` be lent out of the Library under any circumstances. = NorTE.— Certain other works are included in this prohibition, such as costly illustrated works, and volumes belonging to sets which could not be replaced if lost. The GENERAL INDEX to the first twenty volumes of the Journal (Botany) and the Botanical portions of the Proceedings from November 1838 to June 1886 is now ready for delivery to the Fellows on Application. It may be had in cloth, or in sheets for binding. Fellows are desired, when applying for copies, to state in which form they desire to receive it. NOTICE. 3 Vols. XXVI, XXVIL, and XXVIII. are in course of simul- | ‘taneous issue, to expedite the publication of papers, and the Parts already published are as follows :— Yor. XXVI Nos. 178-175. ~~~ (Nos. 176-180 inclusive are reserved for the completion of Messrs. Forbes and Hemsley’s “Index Flore Binensis. 7) > XXVII Nos. 181-188. (Completed with this Part.) i. XXVIII. Nos. 189-192. SCH 1 wilt be issued in a few days. ) s ^ u The Meetings of the Society during the remaining partof the SEH will be as under :— DENM May. Thursday, the 7th. e = aS » Monday, the 25th. (Aaniversary.) ; ES Juse.. Thursday, the Ath. i » » 18th. hs the hour will be Ss Pa. RULES FOR BORROWING BOOKS FROM THE LIBRARY. As amended by the Council, 15th March, 1888. l. No more than Six volumes shall be lent to: one përson - at the same time without the special leave of E Council ` . or one of the Secretaries. : s 2. All books shall be returned before the iust of Six weeks from the time of their being taken out, but if. not ` required by any other Fellow, they may, on M puris be. kept for a further period of Six weeks. 8. All books lent shall be regularly entered LA the Librarian in a book appropriated for that purpose. ot 4. No work forming part of Linnseus" s own Library shall be lent out of the Library under any circumstances. - E Norz.— Certain other works are included in this prese 3 such as costly illustrated works, and volumes belonging to” sets which could not be replaced if lost. ^ i The GENERAL INDEX to the first twenty volumes of the Journal (Borasx) and the Botanical portions of the Poetis from November 1888 to June 1886 is now ready for delivery to the Fellows on Application. It may be had in cloth, or in sheets for binding. Fellows are ce desired, when applying ! ud ome to state in whieh form hey nd. ^ NOTICE. - Vol. XXV. having been completed by the publication of the aat part issued, No. 172 (with Titlepage and Index), and d XXVI. being reserved for the completion of the “Index Flore Sinensis,” of which the last part issued is numbered 173, ue Ne present part, No. 181, commences Vol. XXVII. ‘ ‘The Plates (V.-VIL) to accompany Mr. Massee’s paper on Thelophorea, commenced in this number, will appear with the concluding portion of it, which is in type for early issue. - -to prevent application for unpublished numbers. The. ‘ning Meetings of the present Session will be held as 3? LO 1890. Thursday, April — 38 1890. Saturday | way 24 E m i (Anniversary) y » » 7 Thursday, June 5 » May 1 19 is "The Chair will be taken at Benz o "clock in the Evening pre- cisely at every Meeting, excepting on the 24th May, the day &ppointed for the Anniversary Elections, when the Chair will be at Tumie o "clock in the afternoon. epu undc Ee Attention to this announcement is specially requested, .-