In dense woods on the top of the high hill east of 

 Coonoor, Neilgherries, flowering April and May. 



Blume, the founder of this genus, defines it " mon- 

 cecious, calyx deeply six-parted; lobes in 2 series, 

 petals alternate, shorter than the calyx ; inserted, in 

 the male, round an emarginate stameniferous disk, 

 Male, stamens six, subulate ; cells of the anthers 

 roundish, extrorse ; 3 rudimentary styles in the centre. 

 Female, ovary 3-celled, cells 2-ovuled ; styles three, 

 short, divaricated. Fruit capsular, 3-coccous, with 

 l-2-seed in each. A shrub about 15 feet high with al- 

 ternate, 2-stipulate, elliptic, oblong, entire, coriaceous, 

 glabrous, veined, leaves. Flowers axillary, glomerate, 

 bracteate ; males sub-sessile ; females longish pe- 

 duncled. 



In all these particulars my plant, with the excep- 

 tion of the number of parts of the flower, accurately 

 agi'ees — 5 in mine 6 in his — and as the number may 

 vary, I see no reason, on that account, for forming a 

 separate genus for mine. 



There is however one very important point in mine, 

 to which he does not advert and which merits parti- 

 cular notice as it may yet lead to their separation ; I 

 allude to the structure of the seed. In mine they are 

 exalbuminous ! If in the Java plant they are albu- 

 minous, then that character, added to the difference 

 in the number of the parts of the flower, will claim 

 for the Indian plant a separate generic name. And 

 on the supposition that so accurate an observer as 

 Blume could scarcely have overlooked a circumstance, 

 so rare, in the order, I had in the first instance con- 

 structed a generic character for this plant, under the 

 name of Sarcospemum^ — in aUusion to the structure of 

 the seed — from which I quote the following sentences, 

 " capsule 3-seeded by abortion, seed large, fleshy, 

 exalbuminous: cotyledons unequal, the larger ex- 

 terior one nearly inclosing, and in gi^eat pai't conceal- 

 ing, the interior smaller one." The figures 11, 12, 

 13, and 14 imperfectly represent this formation. With 

 these notes I leave the future disposal of this plant 

 for the decision of observers who may have an op- 

 portunity of examinmg the Java plant. 



1911. Amanoa Indica (R. W.), anthers innate. 

 Courtallum, in alpine jungle. 



Shrubby or sub-arboreous, ramous : leaves altei'nate, 

 oblong elliptic, entire, acuminate, coriaceous, glabrous. 

 Flowers axillary, glomerate, male and female mixed, 

 bracteate : bracts cUiate. Male : calyx 5-parted, lobes 

 ovate with 5-alternate glands adhering to the margin 

 of a glandular disk : stamens 5, inserted round the 

 base of a rudimentary 3-lobed pistil, anthers innate, 

 cells divaricating at the base. Female : calyx, glands, 

 and disk as in the male ; no rudimentary stamens : 

 ovary nearly concealed within the connivant disk, 

 very hairy, 3-celled, with 2-ovulc3 in each : styles 3, 

 deeply cleft; lobes stigmatose: capsule 3-celled, 3- 

 furrowed, obsoletely 3-angIed: cells 1-seeded by abor- 

 tion, seed -. In my specimens none of the seed 



are sufficiently advanced for dissection. 



Of this genus, up to the present time, only one spe- 

 cies has been published, viz. A. Guianemis^ but A. 

 de Jussieu states that he saw 2 others from the same 

 country- The Indian plant differs from his generic 

 character in regard to the stamens, in his the anthers 

 are aduate to the dilated apex of the filament, and 

 extrorse ; in mine they are innate (attached to the 



room for hesitation in placing it in that genus. My 

 herbarium possesses a second species from Ceylon. 

 The two affording new links connecting these dis- 

 tant floras. 



1912-13. PiERARDEA MACRosTACHYs (W. and A.), 



Males ; spikes fascicled on the naked branches : flowers 

 ternate, short pedicelled ; perianth 4-5-partcd> lobes 

 linear, pubescent on both sides : stamens 8-10, insert- 

 ed round the base of a 2- or 3-lobed rudimentary 

 pistil : female ; racemes fascicled as in the male, much 

 longer ; flowers solitaiy in the axil of each minute 

 bract: perianth 5-parted, pubescent: ovary hairy, 

 truncate at the apex, 3-celled : cells 2-ovuled : fruit 

 pulpy, baccate, red when ripe, about the size of a 

 large sti'awberry, 3-celled, 3-seeded: seed compressed, 

 covered with fibrous membrane : no aril : embryo 

 thin, enclosed in copious albumen : cotyledons folia- 

 ceous, orbicular : radicle short, superior. 



Mountain forests, Malabar, Anamallay forests. 

 Western slopes of the Neilghemes below Sisparah ? 



In the above extended, descriptive character, I have 

 felt myself under the necessity of avoiding reference 

 to the leaves, from finding a marked discrepancy be- 

 tween those given on the two plates which I had not 

 observed when preparing them. The leaves shown 

 in 1912 are unquestionably those of a Pierardia^ but 

 I now find they appertain to what appears a different 

 species fi'om the flowers : while those of 1913, though 

 forming part of the specimen, are yet detached from 

 the flowers. They differ from the other in being 

 opposite, hence a suspicion arises that they do not 

 belong to the tree or indeed to the same genus. If 

 on further investigation it turns out that they really 

 do belong to it, the two species may be defined, as 

 regards each other, in two words, the one, " leaves op- 

 posite," the other, "leaves alternate." This difficulty 

 cannot at the present moment be cleared up, but in 

 the mean time it seems to me they are distinguish- 

 able by the flowers alone, in the one, P. macrostacliys^ 

 the segments of the calyx are linear lanceolate, in 

 the other sub-orbicular ; there are besides points of 



habit easily appreciable to the eye but not easily 

 defined in words. 



At one time I thought I could define them by the 

 relative number of stamens and lobes of the calyx, 

 the numbers being equal in the one, 2 to 1 in the 

 other. This I soon found inapplicable in practice, 

 from finding in both great irregularities- The figures 

 in this species give examples of two flowers showing 

 respectively 4-5 sepals and 8-10 stamens; two or 

 three other variations might have been introduced, 

 such as stamens and sepals equal, stamens, 5-6^7, &c. 

 with 4-5, sepals no uniformity of numbers. The 

 other is similarly irregular, so that so far as I have 

 been able to advance it would appear that positive 

 chai-acters are not readily obtainable from the rela- 

 tive numbers of these two parts, though I certainly 

 think that they may be employed if some latitude 

 were allowed. 



P. macrostachySy lobes of the perianth linear lan- 

 ceolate, acute, hairy: stamens usually twice as many: 

 rudimentary pistil 2-3-lobed. 



P. Courtallensis, lobes of the perianth 4-6, sub- 

 orbicular, blunt, covered with veiy short rigid hairs : 

 stamens about equal in number when five or six- 

 lobed : oftener double when four lobed : rudimentary 



point of the filament). In all other points my plant pistil discoid : leaves alternate, samewhat obovate. 



agrees so well with his chai-acter as leaves me no 



cuniate, ending in a short blunt acumen, entire, glab- 



C 30 ) 



