OF CA8TANOPSI8 



loi 



On this account I refer Maingays specimens (Kew DisL 1457 I) here with some hesitancy 

 Ihis species is allied also to C. tribuhides, A. DC. 



Plate S5B.-C. argyrophyUa, King. 4, flowering-branch ; 5, ripe fruit-*/ natural *ix 



13. Castanopsis aumata, $]><ich. Hist. Nat, Plum. xi. 185. 



Young shoots minutely cinereous-toment. oe ; all other parts, exr.pt the infiof - me. 

 glabrous. Leaves thinly coriaceous, oblong-lanceolate, gradually narrow. I to the mu \ 

 acuminate apex, entire; the base rounded or acute; upper .surface shining, the lower dull, 

 pale; nerves 8 to 12 pairs, thin, but prominent on the lower surfa<v and much curved 

 length of blade 4-5 to 7 in., breadth 1-25 to 225 in.; petiole M in. JU 'e sp, f axil! ^ 

 or terminal, solitary, or sub-paniculate, longer than the 1 ives, erect; the Hon 

 solitary, or in glomeruli. Female spikes a few in the panicle with the male- or in s Derate 

 branches, solitary or terminal; the flowers solitary. Ripe fruit from 1*25 to 175 in i,, 

 diameter, ovoid, slightly flattened on one side, thick-walled, externally ciucreous-pubescen 



t 



i> 



and bearing numerous groups of spines arranged in imperfect curving zones; the spin 

 simple or branching, stout, radiating, flattened and pubescent below, but with shaiy, 

 glabrous, apices, from -15 in. to -25 in. long. Nut solitary, compressed-ovoid.— 11 ook. Jit, 



Fl. Br. Ind. v. 622; Miq. Ann. Mus. Lugd. Bat i. 119.— Querent armata, Roxb. PI. Corom 

 iii. t. 296; Fl. Ind. iii. 640; Wight Ic. 770.— Castanea tribuhndety A. DC, var. armata, 

 Kurz For. Fl. Burm. ii. 480.-^ Castanea Falconeri, Hance Joum. 13ot. for 1875, 367 



Castanea sphwrocarpa, Lindl. in Wall. Cat. 3736, and in PI. As. Rar. ii. 5. 



In the forests at the base of the Sikkim and Bhotan Himahn i, in Assam and 

 Cachar, and on the Khasia and Chittagong and tturinah Hill Ranges, up to eh vat ions 

 of 2,000 to 3,000 feet. 



The involucre of this is very characteristic. Its walls are not quite covered, as 

 in C. tribuloides, by the prickles, but these are seen to be arranged in two series of 

 curved semi-zones on the rounded surface, the flattened part being almost free fi 



ro n i 



is 



prickles. In foliage this resembles C. argent ca. but the involucre of the latter 

 almost completely covered by long slender prickles. This species is very badly repre- 

 sented in collections, and this fact no doubt accounts for M. De Candolle having reduced 

 it to O. tribuloides, from which it is really very distinct. Mr. Gustav Mann, Conservator 

 of Forests in Assam, has recently sent numerous specimens from the Assam Hills, while 

 a collector sent from the Botanic Garden, Calcutta, has supplied an abundance of good 

 specimens from the Chittagong Hill Tracts. In Don's Prodromus of the Flora of Nepal 

 Q. armata, Roxb. is quoted as a synonym of Q. tribuloidcs, instead of Q. ferox, the allied 

 species with armed involucres which Roxburgh described from the ime localities as 

 this. I reduce Hance's C. Falconeri to this species without any hesitation. In originally 

 describing C. Falconeri, Dr. Hance remarks that he had never seen a sj 'cimen of Roxburgh's 

 C. armata, but that, judging from his figure of it published by Wight (Ic. 770), the two 

 must be closely allied. Dr. Hance further remarks that both species have very sinuate 

 cotyledons — a character not hitherto recorded in the group, except in Wight's copy of 

 Roxburgh's figure just quoted. From this peculiarity in the cotyledons, Dr. J lance argues 



that M. De Candolle must be wrong in reducing Q. armata, Roxb., to C. trii>uloidet t A. 

 j)Q # — an opinion at which I have also arrived for a different reason. There is no doubt 

 the cotyledons are sinuate in this species. But a character derived from the cotyledons 

 in this family is usually impossible of verification in the herbarium; for, unless the fruit 

 has been thoroughly ripe when collected, its contents speedily decay. I have not therefore 

 thought it worth while to work it. 



