are exactly like the leaf-lobes in miniature." The specimen came from Herb. Lambert, and is labelled in Pavon's hand, "Gossypium N.E."-ED. JOURN. BOT.]

66

Gossypium lanceæforme nob.-Annuum? ramosum ramulis subteneribus, obtuse 4-gonis glabris epunctatis; foliis parum cordatis auriculis minimis supra petiolum imbricatis profundissime inciso 3-lobatis, lobis lanceolatis apice longe attenuatis valde divaricatis, terminali lateralibus 2-plo longiore, integris, e basi 3-5 nerviis, nervis eglandulosis, utrinque opacis et obsolete puberulis in nervis tomentosis, petiolo subpatente, debili striato, subtomentoso limbo dimidio breviore; stipulis sub parvis, lineariacutis, puberulis; pedicello 1-floro oppositifolio petiolo breviore, tereti cinereo-tomentoso; involucro subparvo 3-secto, lobissimo connatis, lanceolato-oblongis, acuminatis, integris, erectiusculis, parallele nervoso glabro; petalis 5 cuneato-rotundatis contortim imbricatis, patentissimis, glabris, flavidis, minute glandulosopunctatis, infra medium macula elongata rubello signatis; tubo staminea petalis multo breviore, undique filamentis brevissimis numerosissimis instructo, antheris subpeltatis, cæteris ignotis. "In Mexico. v. s. in hb. Mus. Brit. (Pavon).”

A NEW SPANISH CERASTIUM.

By A. E. LOMAX.

DURING a botanical expedition in the Sierra de Guadarrama, last June, I discovered a species of Cerastium which does not agree with any description in Willkomm & Lange's Flora Hispanica; it seems to me to be intermediate between Cerastium Gayanum Boiss. and C. Riai Desm., and I propose to name it C. carpetanum.

Cerastium carpetanum mihi. Annua, dense glandulosopubescens, viscosissima; caule a basi divaricato- et dichotomoramoso, 3-6" 1.; foliis sessilibus, oblongo-ovatis vel oblongolanceolatis, obtusis; cymis dichotomis laxi-interdum densifloris; bracteis omnino herbaceis; pedicellis sub anthesin plus minus. curvatis, post anthesin rectis, reflexis, fructiferis demum iterum erectis; calycibus basi subumbilicatis; sepalis altero oblongo ovato vel ovato-lanceolato, obtuso, late scarioso, altero lanceolato, acutiusculo, vix scarioso vel omnino herbaceo, 21" 1.; petalis calyci subæquantibus, vel paulo superantibus, breviter bifidis; staminibus 10; capsula calyce subduplo longiore, 5-61" 1., basi subinflata, apice curvata, attenuata; seminibus reniformis, dorso canaliculatis, bicarinatis, acute striato-tuberculatis, pallide ferrugineis.

In silva in summo jugo supra Puerto de Navacerrada, montibus Carpetanis, Castella, Hispana. Junio.

Between C. Gayanum Boiss. and C. Riai Desm. Differs from C. Gayanum in the shortly bifid petals, obtuse leaves, and reniform, acutely tubercled seeds; and from C. liei in the petals equalling or exceeding the calyx, and the broadly scarious-margined, obtuse sepals.

332

PYROLA ROTUNDIFOLIA AND ITS EUROPEAN FORMS. BY ARTHUR BENNETT, F.L.S.

THIS summer, Mr. Marquand was kind enough to send me some fresh specimens of Pyrola rotundifolia L. from Guernsey. The examination of these in the light of one observation made by Dr. Boswell (Syme) (Bot. Ex. Club Report for 1881, p. 53) on specimens sent by Mr. Sunderland from "The Grande Mare" as "? arenaria -"arenaria I think, J. T. Boswell,"-led me to look up the references to the Lancashire, Scottish, and other plants that have passed under various names in our Floras. So far as I can find, no other botanist seems to have considered the Grande Mare plant other than rotundifolia.

166

The Lancashire plant seems always to have been considered a variety since 1846, as on the 12th November in that year Mr. Kenyon exhibited specimens of that plant at the Edinburgh Botanical Society, and then proposed to call it "P. maritima," giving the differences for rotundifolia. In the 2nd ed. of his Manual, Prof. Babington notices the plant. In the 6th ed. of the British Flora, Hooker and Arnott gave it the name of B. bracteata. In the Phytologist for 1853, p. 1119, Mr. (now Prof.) Öliver called attention to a paper by Planchon in the Annales des Sciences Naturelles (Ser. 3, xviii. 379), where Planchon identifies the plant of Kenyon with the var. arenaria of Koch. In English Botany, Dr. Boswell calls it ß. arenaria Koch, and both Prof. Babington (ed. 8) and Sir J. D. Hooker (Students' Flora, ed. 3) call it by the same name. But Nyman seemed to consider there were two plants under these names, as he has "P. serotina Mlcq. = v. arenaria"; and gives "P. maritima Kenyon (Angl., Belg.)," and localises the v. arenaria from " Scot. Gall. Batav. Ins. Nordern." I have seen

no suggestion in any British Flora as to the Scottish plant, but in Linnæa for 1856, pp. 1-88, Dr. Alefeld gives an account of the Pyrolea, where he mentions that he has seen eight specimens in herb. Hooker, from Scotland, which he makes into a species (under a new genus), as Thelaia intermedia Alefeld, and puts under his plant the var. arenaria of Koch. After this was written, he seems to have come across some mention of Kenyon's maritima, as at p. 88 he names it, and supposes it may be his intermedia.

I have applied to several Scottish botanists for specimens of rotundifolia from near the sea, but I can hear of none such, neither do any of the Botanical Guides suggest such stations.

Another plant was sent to Sir W. J. Hooker (Phytologist, 1. c.), "gathered on the Yorkshire coast, and since found on the shores of Lancashire by Kenyon." This was a mistake, as the locality, which was really on the Durham coast, in a dene running down to the sea, as Mr. J. G. Baker has kindly informed me. Mr. Baker thinks these specimens "approximate a little towards the var. arenaria." Mr. Baker also mentions another station "on the coast near Horden Hill," and wishes, as I do, that someone would look this up again.

We want to know

This is conflicting and unsatisfactory. whether these British plants are all the same thing; are they different from one another? are they really distinct from rotundifolia or are they (with others elsewhere) merely a chain of intermediates? Anyone who has examined a large series of rotundifolia from all parts of the world will be prepared for great variation in the height of the plant, size of the leaves (6"" in diameter! to 36""'!), their shape at the base (tapering to cordate), and the size of the flowers. Of course specimens from the shores of the arctic seas and specimens from damp woods will show great differences; the size of the flower in the former (var. grandiflora) being much larger, at the expense of the leaves, &c.-a usual state in arctic plants.

Dried specimens are not good material to deal with here; given fairly flowered examples of P. rotundifolia, media, and minor, they are easy to separate, but to separate forms of one species is not easy. So I carefully examined five specimens of the Grande Mare plant, and append the result, marking out those characters which are lost in drying:-Style rosy-purplish, shading into purple just below the stigma; stigma deep purple. Anthers yellow to orangeyellow; filaments white. Sepals subparallel for half their length; in many they are fringed at the apex, or slightly jagged; others are subentire, yellowish white (contrasting in this with the much purer white of the petals), paler than as figured in Eng. Botany. Pedicels a little longer than the calyx, when fully expanded. Bracts on the scapes in four specimens, four; in one specimen, three; those on the raceme, eleven, not confined to under the flowers. In bud the apex of the calyx-segments are contracted and recurved, and look nearly entire and subobtuse. The filaments, stamens, and styles are what is called "drusy" in mineralogy; i. e., in white crystallike papillæ, which extend to the inner surface of the petals.

Now this, so far as one can contrast it with dried specimens, is pretty fairly intermediate between rotundifolia and the var. arenaria, perhaps on the whole bearing towards the first. I have not seen a specimen named by Dr. Alefeld, but, looking to his drawing of his intermedia, it seems to me that it is not exactly this Guernsey plant; and he notes under rotundifolia that he has seen three from Guernsey. And I think I am right in saying there is only the one station known in Guernsey; anyhow, Mr. Marquand (Flora of Guernsey, 1891) gives no other.

Yet it would appear that Sir W. J. Hooker did see something in these Scotch specimens that looked different from typical rotundifolia, as Dr. Alefeld records that these plants appeared in herb. Hooker as "P. rotundifolia var. squamosa Hook. MS." Although I have looked through all the rotundifolia at Kew, I did not examine these particularly, as I want to see similar ones in the fresh state. I have seen or possess specimens of the arenaria from all the recorded stations, but here again I want to see fresh Lancashire specimens.

I trust that any botanist who has the opportunity of gathering any of the British plants I have named will carefully examine

(and record) them fresh, and I should be very glad to see them myself.

I have not yet been able to see a specimen of the P. serotina Mlcq., but if that belongs to arenaria, we may put the following names more or less to these rotundifolia forms:

Pyrola rotundifolia Linn. Sp. Pl. ed. 1, vol. i. p. 396 (1753).
B. arenaria Koch, Syn. Fl. Germ. et Helv. ed. 1, p. 478 (1837).

B. bracteata Hook. & Arnott, Brit. Fl. ed 6, p. 276 (1850).

P. maritima Kenyon, Phyt. vol. ii. p. 727 (1846).

P. arenaria Dum. Boug. Lit. Belg. p. 41 (1869).

Thelaia intermedia Alefeld in Linnæa, vol. xxviii. p. 65 (1856).

And according to Dr. Alefeld

“P. rotundifolia var. squamosa Hook. MS.

P. rotundifolia var. albiflora Karel. & Kir. MS.

P. intermedia Schleich. Catalog."

P. serotina Mélicocq, Pl. spont. Béthune in Caffin, Annuaires du Pas-de-Calais, p. 223 (1849); Puel. & Maile, Herb. loc. de France, No. 158, Feb. (1854).

In a subsequent communication to the Botanical Society of France (Bull. Soc. Fr. i. 162 (1854), Baron Mélicocq records that he considers his plant only a variety of P. rotundifolia.

With regard to the eight specimens in herb. Hooker, they are very extreme examples of the Southport plant, and have bracts on the scapes varying from seven to eleven, few flowers, and small leaves; but almost every form between this and the French specimens is represented from Lancashire, although the characters of the pedicels and sepals are better maintained.

A somewhat similar variation occurs in P. minor, on the W. European coast; it is the P. minor var. arenaria Lantzius-Béninga, Beitr. Kennt. Fl. Ostfrieslands, p. 40 (1849). I understand that P. minor occurs in sand near the sea on the west coast of Scotland, but I have not seen specimens.

SYNOPSIS OF GENERA AND SPECIES OF MALVEÆ. BY EDMUND G. BAKER, F.L.S.

(Concluded from p. 273.)

**Folia suprema lobata.
Mexicanum.

137. A. TRILOBATUM Hemsl. Diag. Pl. Nov. p. 24.
Hab. Central Mexico, Parry & Palmer, No. 81!

Austro-Americana.

138. A. MUELleri FredericI Gürcke et K. Schum. l. c.
Hab. Brazil. Prov. St. Catherina, Schenck, No. 497.
139. A. SENILE K. Schum. 1. c. p. 424.
Hab. South Brazil, Glaziou, No. 12438!

140. A. PEDRE BRANCE K. Schum. 1. c. p. 425. Hab. Brazil. Prov. Minas Geraes.

Sida elegans

141. A. ELEGANS St. Hil. Fl. Bras. Mer. i. p. 207. Dietr. Synop. iv. p. 852. S. bella Steud. Nom. ii. p. 576.

Hab. Brazil.

142. A. SELLOWIANUM Regel in Ann. Sc. Nat. ser. 4, xii. p. 379. Sida Sellowiana Kl. in Otto & Dietr. Allg. Gartenzeit. 1836 p. 9. Hab. South Brazil, Glaziou, No. 1457.

143. A. STRIATUM Dicks. in Lindl. Bot. Reg. 1839, Misc. p. 39. A. pictum Walp. Rep. i. p. 324. Sida picta Gill. in Hook. & Arn. Bot. Misc. iii. p. 155. S. striata Dietr. Syn. iv. p. 852; Bot. Mag. t. 3840.

Hab. Brazil. Organ Mts.! Uruguay, Tweedie! &c.

A. Thompsoni Hort. is closely allied to the above.

144. A. NIVEUM Gris. Pl. Lorentz. p. 44.

Hab. Argentine Republic, Lorentz, No. 175; Hieronymus & Lorentz, No. 922.

145. A. REGNELLII Miq. in Linnæa, xxii. p. 554. A. septemlobum Miq. l. c.

Hab. Brazil. Prov. Minas Geraes. St. Paulo. Rio de Janeiro, Glaziou, No. 18891!

146. A. DARWINII Hook. in Bot. Mag. t. 5917. A. Hildebrandtii Fenzl in hort.

Hab. Brazil. Prov. St. Catherina !

Var. TYPICA Regel in Gartenflora, xxv. p. 317.

Hab. Brazil. Prov. St. Catherina.

Var. TRINERVE Regel, l. c. xxiii. p. 130, t. 794.
Hab. Brazil. Prov. St. Catherina.

Var. EXPANSUM Regel, l. c. xxv. p. 317.

Hab. Brazil. Prov. St. Catherina.

147. A. VENOSUM Walp. Ann. ii. p. 158; K. Schum. I. c. p. 431,

t. 76. Sida venosa Hort. in Bot. Mag. t. 4463. He

Var. B. BREVICALYX K. Schum. 1. c, p. 431.

Hab. Brazil. Prov. St. Paulo. St. Catherina.

Var. y. LANATUM K. Schum. l. c. p. 431.

Hab. South Brazil, Mendonca, No. 1050.

Sect. II. CORYNABUTILON K. Schum. l. c. p. 369. Stigmata decur

rente papillosa.

*Folia parva.

148. A. BICOLOR Phil. in Anal. Univ. lxxxii. p. 322.; K. Schum.

149. A. CERATOCARPUM Gay, Fl. Chil. i. p. 331. Sida ceratocarpum Hook. & Arn. Bot. Misc. iii. p. 154. S. stelligera Poepp. Coll. Pl. Chil. iii. No. 172.

Hab. Chili. Santiago! Campana di Quillota!