29. Phellopterus litoralis, Benth. Ad litora marina, juxta Chi fu, coll. W. Hancock. Not previously found in China.

30. HERACLEUM MELLENDORFFII, sp. nov. Caule plurisulcato aspero, foliis sparsim brevissimeque hispidis trisectis (summis tantum trilobatis) segmentis longe petiolulatis basi cordatis ad medium palmatifidis grosse inciso-serratis, umbellis radiantibus, fructibus ellipticis utrinque obtusis spinulis minutis sparsis hispidulis 3 lin. longis vittis dorsalibus lateralibusque lineari-subclavatis una alterave sæpissime deficiente commissuralibus 2 clavatis parallelis omnibus medium mericarpium adtingentibus v. paulo ultra pratensis. In m. Po hua shan, China bor. d. 6, Octobris, 1874, legit Dr. O. von Moellendorff. (Herb. propr. v. 19,197).

This is probably the plant referred doubtfully in Maximowicz's 'Index Floræ Pekinensis' to H. dissectum, Ledeb., from which it is apparently quite distinct.

31. Oldenlandia alata, Koen. In rupibus humidis inter silvas ad Tingü shan, prov. Cantonensis, ineunte Junio, 1867, detexit Sampson. The specimens belong strictly to Hedyotis pterita, Bl., distinguished by the capsule with four equally broad wings, which Decaisne (Herb. Timor. Descript.' 89) says is the same as Koenig's species, an opinion the accuracy of which is questioned by Miquel. A character not noticed before, so far as I know, is that the capsule, while quite fresh and succulent, dehisces at the top, and exposes the perfectly white and immature seeds, which are subsequently blackish brown and deeply pitted. I cannot, after examining the living plant, think with Mr. Bentham (Fl. Hongkong.' 152) that this is reducible to O. racemosa, Lam.

32. Cyathocline lyrata, Cass. In prov. Cantonensi, vere 1866. T. Sampson.

33. Antennaria leontopodina, DC.? I have received this from a number of localities in North China, where it seems very common. M. Maximowicz assures me that it is the Leontopodium sibiricum 7. depauperatum of Turczaninow, and a reference to vol. ii. Add. p. 38, of that author's Flora Baicalensi-Dahurica' will show that he proposed, doubtfully, naming it A. Steetziana. The limits between many Helichryseous genera seem to me altogether fanciful; but I do not believe the present plant can by any careful observer be united with Leontopodium sibiricum, Cass.

34. Anaphalis triplinervis, Benth. In m. Siao wa tai shan, legit Hancock. New to China. The leaves of the Chinese plant are very much smoother than in Blinkworth's Kumaon specimens, but I do not think it can be separated specifically.

35. Microrhynchus (Launaa) sarmentosus, DC. Longissime repens et radicans in arena litorea profunda ac mobili ad Pak sha, extremitate australiori prov. Cantonensis, ex adverso ins. Hai nan : ipse legi d. 20 Novembris, 1866.

36. Pieris (Eupieris) lanceolata, Don. Inter saxa in summo cacumine montium Pakwan, supra Cantonem, d. 26 Martii, 1870, leg. Sampson. Only previously recorded from the mountains of India.

37. Stimpsonia chamaedryoides, A. Gray. Prope Fuchau, Maio,

1857, legit Gregory. Japan is alone mentioned as the native country in the Genera;' but I sent one of Gregory's specimens to Kew, when first detected.

38. Mitrasacme indica, Wight. Ad Swatow, in scaturiginosis, duo tantum exemplaria pusilla invenit Sampson, Junio, 1866. Circa Amoy ipse legi, Oct. 1857. Found hitherto only in the Indian peninsula, Ceylon and Australia.

39. Calotropis gigantea, R. Br. In parva insula Nau chau, juxta Hainan, m. Julio, 1870, leg. Sampson.

40. Lithospermum Zollingeri, A. DC. In monte Feng wang shau, circ. 35 mill. pass. ab urbe Shanghae, æstate 1876, coll. J. P. Martin. Only previously gathered in Japan. Maximowicz, who no doubt rightly reduces to this L. japonicum, A. Gray, refers it to the section Margarospermum ( Mél. Biolog. Acad. Pétersb.' viii. 542), next L. purpureo-caruleum, L. Lithodora is not coextensive with this, as stated by Willkomm (Prodr. Fl. Hispan.' ii. 499) and Bentham and Hooker (Gen. Plant.' ii. 860); for Reichenbach includes in his section L. purpureo-cæruleum, which Grisebach (Spicil. Fl. Rumel.' ii. 85) expressly excludes, and I think he and Boissier (Fl. Orient.' iv. 218) right in associating this species with L. officinale. As remarked by Bentham, L. apulum, Vahl., and its ally L. microspermum, Boiss. (= L. Webbii, Coss. & DR.) form a good section of themselves, for which Columna's name Apula might well be restored.

41. Ipomea (Euipomaa) Turpethum, R. Br. Prope Whampoa, ipse legi, d. 16 Februarii, 1871; secus amnem Lien chan, Oct. 1876; coll. Rev. J. C. Nevin.

42. Ipomea (Pharbitis) congesta, R. Br. In insula Prata, maris chinensis, m. April, 1858, leg. C. Wilford. The peduncles are 1-flowered, and the corolla rather more than three inches in length in my specimens.

43. Pedicularis longiflora, Rudolphi. In m. Siao-uri-tai-shan, Julio 1876, coll. W. Hancock. This curious and very distinct species had only hitherto been found in Baikal, Siberia.

44. Thunbergia fragrans, Roxb. Serpens inter gramina secus viam a portu Hoihan ad metropolin Kieng-chan-fú ins. Hainan ducentem, d. 18, Novembris, 1866, coll. Sampson et Hance.

45. Dadalacanthus nervosus, T. And. In umbrosis riparum scopulosarum præcipitum ad fauces Shiuhing, secus fl. West River, Febr. 1857, legit Sampson. Not previously recorded from China.

46. Avicennia officinalis, L. In ins. Kulangsu, Amoy, Julio, 1865, coll. Sampson.

47. Polygonum multiflorum, Thunb. Circa Chinkiang, 1876, coll. Stronach. A rare Chinese plant, as far as I can judge from different envois.

48. WIKSTREMIA MONNULA, sp. nov. Ramis angulatis glaberrimis atropurpureis, ramulis pubescentibus, foliis oppositis v. alternis herbaceis ellipticis acutis supra glaberrimis opacis subtus pallidis tenuiter venosis sparsim pilosulis 9 lin. longis 4 lin. latis petiolo lineali, racemis ramulos terminantibus capituliformibus

8-12-floris, perigonio gracili sericeo-pubescente in sicco luteopurpurascenti semipollicari lobis brevibus obtusiusculis tubo triplo brevioribus, antheris superioribus subexsertis inferioribus medio. tubo sitis, ovario oblongo apice sericeo-pubescente, stylo brevi glaberrimo stigmate magno capitato, squamulis hypogynis binis lineari-oblongis viridibus ovario triplo brevioribus. In prov. Cantonensi, secus fl. North River, m. Martio, 1877, coll. T. L. Bullock. (Herb. propr. n. 19,989.)

A pretty and delicate species, apparently quite distinct from any yet described.

Caule

49. ARGYROTHAMNIA (Speranskia) CANTONENSIS, sp. nov. tomentoso, foliis alternis flaccidis ovato-oblongis utrinque obtusis grosse et inæqualiter crenato-serratis serraturis sinubusque minute glandulosis supra sparsim pilosis subtus tomentosis penninerviis costa nervisque subtus prominulis copiose et minute pellucidopunctatis 1-2 poll. longis 7-9 lin. latis petiolo 3-9 lineali, stipulis nullis, racemis terminalibus hirsutis, floribus binis v. sæpius ternis pedicellatis nunc in ramulos breves dispositis unisexualibus 1 lin. diametro bracteis parvis lanceolatis fultis, sepalis extus hirsutis lanceolatis, petalis tenerrimis albidis deltoideo-obovatis unguiculo brevi sepalis paulo brevioribus, staminibus 10-15 triverticillatis parum exsertis filamentis glaberrimis, ovario tridymo hirsuto tuberculis grossis oblongis obsito, stylis hirsutis bifidis stigmatibus laceris. Secus fl. North River, prov. Cantonensis, Martio, 1877, coll. T. L. Bullock. (Herb. propr. n. 19,955.)

An interesting plant, resembling in general appearance its near relative A. tuberculata, Müll. Arg.! from North China, but differing by its long-stalked proportionally wider leaves, with crenate teeth, the entire absence of stipules, and the smaller flowers and bracts.

50. Euphorbia Esula, L.; E. cyparissivides, Boiss. Secus fl. North River, Mart. 1877, leg. Bullock. Common apparently in the extreme north of China, but not previously recorded from the south of the Empire.

51. Chloranthus japonicus, Siebold. Secus fl. North River, Martio, 1877, leg. Bullock. A singularly meridianal station for a plant heretofore detected only in Japan, the Korean archipelago and Machuria.

52. Trewia nudiflora, Willd. In fruticetis juxta Cantonem, Aprili, 1870, detexit T. Sampson. New to China.

53. Eriocaulon echinulatum, Mart. In uliginosis extra Cantonem, secus viam ad montes Pakwan ducentem, d. 16 Octobris, 1866, detexit T. Sampson. This pretty little species had hitherto only been recorded from Burma.

54. Cyperus Wightii, N. ab E. Whampoa, Maio, 1875. The Chinese plant is quite identical with an Assam specimen gathered by Jenkins, for which I am indebted to Dr. George King, who gives C. Zollingeri, Steud. as a synonym. C. compressus, var. spiculis angustis, Thw. (C. P. 807) is also undistinguishable, and Dr. Harland gathered the same plant in Singapore. His specimen in my herbarium was doubtfully identified by Mr. Bentham with C.

lucidulus, Klein. Although referred to C. compressus by Thwaites, Nees and Steudel both place it in a separate paragraph of the genus. This species is new to the Chinese flora.

55. Polypodium hirtellum, Bl. In summo monte Tai-mo-shan, i.e., 'mons ingens molarum'e regione ins. Hongkong, alt. 3000 ped., d. 23 Februarii, 1877, coll. T. Sampson. Found hitherto only in Java and Ceylon. The Ceylon fern (C. P. 3902) given me by Dr. Thwaites as named by Sir William Hooker himself P. lasiosorum, is certainly not distinct, nor has it the long stalk figured by Blume (Fl. Jav. Fil.' t. 46, f. 6) on which its distinction was mainly based. Thwaites' C. P. 3921, P. parasiticum, var. latiusculum (not pilosiusculum, as misnamed by Mr. Wall, 'Cat. of Ceylon Ferns,' p. 7) appears to me distinct from the species to which it is referred, and nearer the present one.

56. Polypodium lineare, Thunb. In m. Siao-wu-tai-shan, Jul., 1876, coll. W. Hancock. Although occurring in Japan, and at various points along the Chinese coast, this is, I believe, the most northerly point on the Asiatic mainland where it has been hitherto gathered. The Russian writers and Milde do not mention it as occurring in the Peking district, Manchuria, or Siberia.

57. Cystopteris montana, Link. In m. Siao-wu-tai-shan. W. Hancock. This is the first record of this fern from Eastern Asia, so far as I am aware. The specimens agree perfectly with

Norwegian ones.

Cum priore.

58. Cystopteris fragilis, Bernh. 59. Woodsia glabella, R. Br. Cum duabus præcedentibus. Now first recorded from China, but it had been previously found in Dahuria, Siberia, and Sachalin.

60. Hymenophyllum (§ Glabra*) emersum, Baker. In rupibus summi montis Tai-mo-shan (mons ingens molarum) adversus Hongkong, alt. 3000 ped., d. 23 Februarii, 1877, coll. T. Sampson. Only known previously from Ceylon and Mauritius.

NOTES ON SOME SOUTH-EAST DEVON PLANTS. BY THE REV. W. MOYLE ROGERS.

I.-INLAND.

ALL the plants recorded here have been observed by me in the course of the last twelve months in the parish of Trusham, or its immediate neighbourhood. Trusham is a small parish of between 700 and 800 acres, lying about midway between Exeter and Dartmoor, on the south-west slope of the Haldon Hills, the crests of which rise between it and the nearest sea-coast, nine miles distant, at Dawlish. It has for its contiguous parishes Hennock and Christowe on the west, with the river Teign as boundary between, Chudleigh on the south and east, and Ashton on the north. The

* Confr. Prantl, Untersuchung. z. Morphologie d. Gefasskryptogamen. I. Hymenophyllaceen, p. 54.

soil (the Ordnance Geological Chart is my authority) is carboniferous, with greenstone cropping up here and there, limestone occasionally to the south-east of Chudleigh, and granite above Hennock; while to the west of Bovey Tracey (a town beyond Hennock, and five miles by road from Trusham) lie the "Bovey Beds" (Lignites and Clays). This last-named district, known in the neighbourhood as the Bovey Heathfield, is said to be rich in interesting plants; but I have as yet visited it only once, and that so early as May. The semi-maritime character of the flora of Trusham parish appears to me to be its most marked special feature, especially when it is remembered that the Haldons intervene between the district and the sea.

Ranunculus parviflorus, Linn. Exceedingly abundant on open down and in stony field throughout the district, and quite a feature in the flora-especially as the foliage keeps freshly green through the greater part of the year. Also very common on the coast of South-east Devon, near Sidmouth, at Budleigh Salterton, and on Dawlish Warren.

Cardamine impatiens, Linn. In "Teign Lane," Trusham, where it was discovered by Mr. T. R. Archer Briggs in May last, while the plants were still only an inch or two high. Long known by the Rev. H. Roberts at Ashton, in a lane parallel to Teign Lane, but from two to three miles to the north of it; and on his authority recorded for South Devon in Topogr. Bot.' In Teign Lane this year it flowered freely, from the middle of June to the end of September, along the dry bushy bank for a quarter of a mile or more.

Barbarea intermedia, Boreau. In Trusham, Ashton, and Christowe parishes; rather frequent in waste spots and in rocky ground, especially by the river Teign; but never many plants together. B. pracox, Brown, is often to be seen near it, but is most abundant in the borders of stony fields.

Teesdalia nudicaulis, Brown. Very common throughout the district in bare rocky ground, and not infrequent in thin copses.

Viola odorata, Linn. (with purple flowers). Very abundant at the Chudleigh Rocks, but occurs nowhere else in the district, so far as I have yet seen. The more common variety (b. alba of Lond. Cat., ed. 7) is fairly general in woods and lanes. The lilac form, which on chalk in Dorset and South Wilts I have always found (without difficulty, though usually in small quantity) growing with the purple and the white, seems quite absent here. The Chudleigh Rocks are also remarkable for the great abundance of V. Reichenbachiana to be found there-almost to the exclusion of the more general V. Riviniana; the exact reverse of this being true at Trusham, Ashton, and apparently elsewhere in the district.

Viola lactea, Sm. Common on the Bovey Heathfield, on Knighton Heath (which is a continuation of Bovey Heath), and on similar ground near. With it, but (so far as I observed) only on the drier and more strictly heathy spots, occurs a form which suppose to be the var. b. intermedia of Lond. Cat.,' ed. 7.

I