304 



The carpels, which are completely consolidated in the flower, 

 remain so until they have attained their final size, which may be 

 as much as 1*5 m. in length and 2 cm. in diameter. They then 

 begin to separate from the base upwards into two follicles, 

 without, however, becoming detached from the torus, and at the 

 same time they dehisce along their ventral sutures. For a long 

 time the carpels remain connected near the apex, but little force 

 is required to cause their complete separation. Most of the seeds 

 probably escape whilst the long, pendulous fruits are in that 

 condition. The structure of the seeds is exactly like that of 

 Alafia landolphioides. At the time of the publication of the 

 Apocynaceae of the Flora of Tropical Africa the embryo of 

 Alafia landolphiensis, and, in fact, of Alafia generally, was 

 unknown. Good seeds, however, have since been received from 

 Southern Nigeria (Unwin, No. 7). The embryo is characterised 



• i^? i lg green colour and bv a longitudinal fold along the 

 middle line of each side-half of the cotyledons, the margins 

 standing out at almost a right angle to the fold.' Exactly the 



ifl 



Holalafi 



are doubled back on the fold. I may add that the seeds of only 

 one other species of Alafia are known, viz.. A. lucida. Stapf 



(Uganda, Why te, without number). Here, too, the embryo is 

 green, but the cotyledons are flat. 



Holalafi 



A lafi 



described, is thus 



IT.JT,-^ rehed upon for its £ ene «c distinction is not 

 correlated with any character alien to Alafia, and the generic dis- 



TTnT^V V! i ardly any lon ? er justifiable. I would, in reducing 

 ^alafia multifiom to Alafia multiflora, not even propose to 



to > innnn" 1 / Bpe ^ su b ; genus or section, so intimate appears to me 



knowK ? + l i' landol P hioi ^ in the light of my present 

 Knowledge of those plants. 



att5n C R ?<£$£ th / d T ription of the seeda > l would add that the y 



Uieir sflivT f 2 ~' ] Cm ' a " d a width of ab °*t 3-4 mm., whilst 

 their silky brown coma measures from 9 to 15 cm. in length. 



Mission i^T'r figU ? d u an old ' Aattened-out follicle in his 

 ad * ^ somf To! i v mrm /' tab - 144 ' and on P- 576 of the same work 

 a 1 wir th^v! 168 , 1101 T ?/ ded so far foi " this secies. They are 

 ciSlWmP n %° n n° f KatU (BOUth of Lake Leopold II.), near 



indicated In th. m ' °?T ^^ and add nothi ^ to tb * area 

 indicated in the Flora of Tropical Africa. 



10 



XLIII.-MISCELLANEOUS NOTES. 



Q^*™^l££l 9 B ' Sc -' FLS ' Director of Public 



F.L.S., Superintendent of thT??' *** MR ' J ° HN H ' HART ' 

 ment, Trinidad ^ h u-p .f- / ?° tame and Agricultural Depart- 

 Fawcett w a i apndnttl f ?i? fr ° m their re «Pective posts. Mr. 

 December! 1886 P aid M r V! ?° 8 f Which he has Just vacated in 

 March, 1887 aft t00k U P his d^es in Trinidad in 



