ROSACEA. 



425 



Acioa guicmensis. 



but the existence of a rudiment of this internal process of the 

 ovarian wall in Couepia proves the close relationsliip of that genus to 

 Parinari and the little value of the latter one.' 



In the Coupis" (figs. 502, 503) we find the receptacle of Hirtella 

 Courpia and most species of Parinari, with its deep unilateral tube. 

 In the fiowers of Acioa ffuianensis, the pro- 

 totype of this genus, we find five imbricated 

 sepals, and as many alternating imbricated 

 petals. There are from ten to fifteen fer- 

 tile stamens all on one side of the flower ; 

 on the other, that of the receptacular 

 tube, the androceal ring only bears a 

 variable number of little sterile teeth. 

 But the filaments of the fertile stamens 

 are united together for a long way up into 

 a long strap-like band, spirally involute 

 in the bud, and then spreading a long 

 way on one side of the expanded flower. 

 Towards the apex the filaments become 

 free and support each a short introrse two- 

 celled anther dehiscing longitudinally. 

 The gynceceum is inserted and formed as 

 in Hirtella. Its very long basilar style 

 ends in a scarcely dilated stigma ; it is 

 rolled up in the bud and exserted on an- 

 thesis. The fruit is a drupe with a usually dry hard thick mesocarp 

 and a one seeded endocarp ; the radicle of the embryo is inferior. 



Fig. 502. 

 Longitudinal section of flower {\). 



^ We have been unable to study the genus 

 Trichocarya (Miq„FL Ind.- Bat., i. 357 ; Suppl., 

 i. 116), which Bentiiam & Hooker phxce 

 {Gen., 607, n. 6) beside the genus Parinari, and 

 whicli, with the characters of tlie last species, a 

 pentamerous flower, an androceum of about 

 twenty-five stamens insei'ted all round the flower, 

 is said to present a tubular, cylindrical receptacle, 

 entirely tilled up by an accrescent gynophore. The 

 ovary, inserted on top of the receptacular tube, 

 is described as containing one (?) ovule, and as 

 accompanied by a basilar style. The fruit is a 

 drupe, with a coriaceous cndocarj) lined with 

 hairs. Two species of this genus liave been de- 

 scribed, one from liorneo, the other from Su- 

 matra,nained by Koutuals {Ned. Kruidk.Arcli., 

 iii. 381, 388) Augeles'a sphndens and Diemenia 



racemosa respectively. Were it not for the sin- 

 gular character of the gynophore, which, accord- 

 ing to MlQUEL himself, needs verification, it" 

 would seem that these plants should be relerred 

 to the genus Parinari. In his last work on these 

 plants MiQUEL {Ann. Mus. Lugd.-Bat., iii. 236) 

 has maintained as distinct genera Angelesia and 

 Diemenia, whicli, as we have just seen, had been 

 united by Bentuam & Hooker in 1865. H. 

 appears that the herbarium at Leyden does not 

 contain enough specimens to allow any clear and 

 definitive judgment to be given on these points. 

 - Acioa Aunx., atdan., 698, t. 280. — DC, 

 Prodr., u. 526. — Spach, Suit, (i Biiffun, i. 371. 

 — H. ^I'i., Adamonia,\i. 222. — Ada (part.) W., 

 loc. cit. — Dulacia Neck., Elem., n. 1236 (c.k 

 DC). — Couepia (part.) I!. H., Gvn., 608, n. 9. 



