Plates 2573 and 2074. 



HEVEJE SPECIERTIM PLURIUM ANALYSES. 



EUPHORBIACE^. 



These plat s are intended to elucidate the floral structure of 

 some of the species of Hevea. The drawings were all made from 

 type specimens, except those representing //. brasiliensis and II. 

 guianensiS) concerning which, however, there is perhaps no doubt, as 

 they are quite distinct from each other, and have been generally 

 recognised. Unfortunately, flowers of the same stage of development 

 could not always be obtained, consequently the conditions for com- 

 parison are not so satisfactory as they might be. Many more flower 

 were examined than were drawn, and the conclusions arrived at may 

 not, in all cases, appear justified by those represented ; but finality is 

 not claimed for these investigations. Mueller laid great stress on the 

 disk, and established his H. janeirensis (Fl. Bras. xi. 2. p. 706) on the 

 assumed absence of disk in the male flowers and some other minor 

 characters. We cannot distinguish it from //. hrasiJien<>s. Thes 

 glands are often exceedingly minute, and easily overlooked in young 

 flowers. The androecium affords much more reliable characters, 

 though they are probably not absolutely constant. Usually the anthers 

 are ten in number, and arranged on a slender column in two contiguous 

 or separated whorls ; but in //. guianensis, Aubl., there are only five, 

 arranged in one whorl. In II. lutea, Muell. Arg., there are from 5 to 

 7 anthers in a broken whorl or whorls. — \Y. BoTTING Hemsley. 



H. brasiliensis, Muell. Arg. in Linncea, xxxiv. p. 204. H. janeiren 

 sis, Muell. Arg. in FL Bras. xi. 2, p. 70G. Siphonia brasiliensis, II. B. K. 

 Nov. Gen. rt Sp. vii. p. 171 ; CoIUtis, Ileport on Caoutchouc, j>\. 1. (1872). 



Brazil : ' Province of Rio de Janeiro,' * Glaziou, 4911 (janeiret >. 



Plate 2.")73 : l*'ig. 1, an expanded female flower and bud of a male flower; 2. pistil. 



* 



Brazil : ' Rio de Janeiro,'* Glazlo)/, 8921 . 



Fig. 3, a male flower; 4, andrcecinm. 



* Aim t certainly an er as there is no other record of anj peeies so far sooth 

 Perhaps cultivated. 



