MR, BEKTHAM ON LOaANIACE-fi. 85 



I have mentioned but two Qardnerias, altkough four are de- 

 ficnbed; but this has been done chiefly on phytogeographical; 

 grounds. On a careful comparison of a considerable number of 

 - specimens, I can find no difference between the Nilgheny G, Wal- 

 itchtana and the original G, ovata from Khasiya ; and a specimen 

 communicated to me by the Leyden Museum of the G. nutans, 

 oieb. et Zucc, from Japan, agrees in every respect with the G. 

 mguBtifolia, which is so abundant in Sikhim, Nepal, and Khasiya. 



W , 



24 Pagamea, -4z^/. 



_ *r - _ X 



The ovary and fruit of this genus have been hitherto but little 

 i^iown. The seeds of Aublet's species, the only one as yet pub- 

 hshed, had not been described at the time I determined Schom- 

 burgk' s LoganiacecB ; and, misled by fruits apparently nearly ripe, 

 but filled with a fleshy pulp which had become granulated by 

 orying, I totally misunderstood their structure. I have conse- 

 quently misled DeCandolle also, who, in the ' Prodromus,' copied 

 the character I had given, in contradiction to the older but more . 

 correct one of Jussieu. Mr. Spruce's labours have now supplied 

 us not only with ripe seeds of the old P. guianensis, but also with 

 fiae specimens, in various states, of three other very distinct new 

 species, from whence I am now enabled to complete tlie history of 

 the genus. 



The flowers, like those of many Bubiaceee, have a tendency to 

 become polygamous, by the abortion of the female organs in some 

 flowers, and occasionaUy of the anthers in others. .The cjilyx, 

 corolla and anthers are correctly described in the * Prodromus,* 

 except that in one species the chaff-like hairs which line the lobes 

 of the coroUa are exceedingly short. The ovary, in many speci- 

 mens of P. guianensis, and occasionally also in the other species, 

 18 short and fleshy, with two very small ceUs containing each a 

 "minute abortive ovule, and the style is then very short, usually 

 •divided to the base. Where the ovary is perfect, which I have 

 seldom had occasion to observe in the common P. guianensis, it is 

 |nuch less fleshy, completely divided into two cells with one ovule 

 jn each, erect from the base, precisely as in Psgclotria. The style 

 IS then elongated, divided much below the middle m P. guianensis, 

 ^ut less so in the others. The fruit is a greenish or black berry, 

 or rather drupe, containing two crustaceous, almost bony pyrenes, 

 ^th a flattish inner face and convex back. Each contains a single 

 erect seed with a thin testa adhering to a cartUaginous albumen, 

 ^ery much ruminated, as in Grumilea. The embryo is small. 



ii 



■ it 





