Guatteria.] flora indica. 139 



oblongus, basi intus sulcatus. Torus parum elevatus, plano-convexus, 

 iuterdum medio excavatus. Carpella sicca, pericarpio tenui ssepe fra- 

 gili. Semen erectum. — Arbores stipe procerce, vel frutices {Inter dam 

 scandentes ?) habit u mrim % foliis oblique nervo&is, iuflorescentia axillari 



vel oppositifolia . 



We retain the genus Guatteria nearly as left by Blume and Martius. It is still 

 very extensive, and perhaps not quite natural, though we have not been able to 

 find any good characters for subdividing it. The greater number of the species are 

 American. 



1. G. longifolia (Wall. Cat. 6442 !) ; foliis e basi lata longissime 

 angustatis, floribus versus apicem pedunculi axillaris racemosis, petalis 

 elongatis, carpellis ovoideis. — W. et A.! Prod. i. 10; Wight, Ic. t. 1. 

 Uvaria longifolia, Lam. ; Roxb.! Fl. hid. ii. 664. Unona longifolia, Dun. 

 Anon. 109 ; DC. Syst. i. 492, Prod. i. 90. 



Hab. In sylvis Zeylaniae ! et Tanjor, Wight! per totaru Indiana 

 tropicam frequentissime culta. — (v. v. cult.) 



Arbor procera, elegans, ramis adscendentibus. Hamuli graciles, glabri ; gemmae 

 pubescentes. Folia basi plerurnque in petiolum angustata, rarius rotundata, sub- 

 membranacea, margine undulata, 5-8 poll, longa vel iuterdum Jfere pedalia, 1-2 

 rarius 3 poll, lata, pet. j-Hhpoll. ; utrinque glaberrima, lucida, pellucido-punctata, 

 nervis obliquis parallclis, venulis crebre reticulatis. Pedunculi ad axillas foliorum 

 delapsorum, breves, £ poll, longi vel plerumque multo breviores, interdum vix ulli, 

 solitarii vel bini (et tunc 1 sessilis), cinereo-incani, basi bracteolis minutis squamse- 

 formibus muniti. Pedicel U plurimi, secus pedunculi apicem dense racemosi, brac- 

 teis parvis suffulti, elongati, graciles, 1-2-pollicares. Sepala extus pubemla. Petal a 

 NpoIL, e basi lata subulata, parallele nervosa. Torus fructus pubescens. Carpella 

 plerumque pauca, f poll, longa, pedicello i-poll., glabra, fusca, vix granulata. 



This well known and very ornamental tree is commonly planted along roads in 

 Bengal and throughout the southern parts of India, but scarcely at all beyond the 

 tropics, the winters of the northern parts of Hiudostau being probably too cold for 

 it. Roxburgh did not know its native country, and wc learn from Blume that it is 

 not a native of Java. It appears, however, to be really indigenous in Ceylon, and 

 in the southern part of the Madras Peninsula. 



2. G. bifaria (Alph. DC. Mem. 41); foliis ellipticis acuminatis 

 basi acutis punctatis superne glabris subtus pubescentibus, pedunculis 

 axillaribus nudis 1-floris, petalis lineari-lanceolatis, carpellis loiige sti- 



pitatis ovoideis.— Wall. Cat. 6447 ! 



Hab. In Ava circa Prome, Wall.! — {v. s. sine fore.) 



This species is said to differ from the next only by the linear petals. The speci- 

 mens in the Linnean Society's herbarium, which are the only ones we have seen, are 

 unfortunately not in flower. We cannot, however, distinguish them in any way from 

 G. cerasoides ; our diagnosis is, therefore, taken verbatim from Alph. De Candolle. 



3. G. cerasoides (Dunal, Anon. 28) ; foliis lanceolatis vel ob- 

 longo-lanceolatis subtus pubescentibus, pedicellis 1-3 ad apicem pe- 

 dunculi axillaris tuberculifornois, petalis ovato-oblongis, carpellis ovoi- 

 deis apiculatis longe pedicellatis.— DC. Syst. i. 503, Prod. i. 93 ; Wall. 

 Cat. 6436 ! W. et A.! Prod. i. 10. Uvaria cerasoides, Roxb.! Cor. i. 



t. 33, Fl. hid. ii. 660. 



Hab. In moutibus tropicis et subtropicis Bahar, HamJ Orissa, 



