1 1- • . 



I 





I* 



222 



MR. L. V. LESTKR-GAllLAXD : A REVISION 



Pflunzonfam. Nachtr. i. 1897, p. 200. 



Thi^ 



S IS 



a f^urprij^inp; mistake, for 



4 - 



7?. polygalacea is not a Delaria at all. l)iit a Bravteolaria. In 1841 , 

 Horhstottor (Flora^ xxiv, p, 058) piiMishoJ ]iis gonus Bradeolarm^ founded on 

 B, racemosa from Natal. Bravleolaria seems to liav(^ boon first roducod to 

 a soction of i?a^;/i/a in Walp. Ann. vil. p. 801 (18G8), The arrann-rment now 

 generally accepted by which Bradeolaria and Delaria are rank(vl ns the two 

 component sections of the genus BapJiia, dates from Bcntham and Hooker, 

 Gen. Plant, i. p. 553 (1875). It may bo remarked, hoAvever, that B. niliJa is 

 there erroneously referred to the section Bradeolaria^ and that the sectional 

 distinction ilrawn from the colour of the flowers (white in Bradeolaria^ 

 yellow in Belana) does not hold for all the species now known. 



The primary division of the genus is based upon the different methods of 

 rupture of the calyx, which before the flowers open forms a sort of calyptra 

 completely enclosing the petals. In Bracteolaria it becomes 2-cleft, ahnost 

 bihibiate, and no subdivision is necessary as the section only includes 

 12 Sj)(^cies. In Delaria^ to which the bulk of the species belongs the calyx 

 splits up one side and forms a sort of spathe. Here it becomes necessary to 

 subdivide. Harms (Pflanzenfam. Nachtr. i. p, 200) took the inflorescence as 

 Ins test, and establislnMl two subdivisions — (a) Gemnna\ with flowers 1 to 5 

 in the axiis of the lcav(^g, which ho further subdivided according to the shape 



I 



illi 



axiiiarv 



and size of the bracteolcs ; and (b) Racemifenc^ with flowers in 

 racemes or panicles. This is not altogether satisfactory, owing io the confused 



J 



cliuraoter of the inflorescence in some species ; and I have substituted an 

 arranjj;einent based on tlie bracteoles, which has the advantat>*e of <dvin<r i ro- 

 minence to a small but very tlistinct group of species, llie Striata^ and seems 



to mc to be on the whole more workabh^ and to throw the species into more 

 natural groups. For fiu'tluu* subdivision I have relied a good deal upon the 

 character of the indumentum of the petioles and inflorescence. 



A third section of the genus, Macrobaplda^ was suggested by Harms in 

 Engl. Jahrb. xl. p. 35 (1908) for his new species, B. macroealf/.v — ''calyx 

 apice distincte 5-dentatus, dentibus majusculis, sub anthesi spathaceo-fissus.'* 

 But this character is obviously of very inferior importance to those used to 

 differentiate the two okhu' sections. Moreover^ the division of the calyx- 

 te(*th is often clearly observable in some other spccdes — indeed, it is a 

 conspicuous feature in Desvaiix's original plate of Delaria pyrifolia. 



BapJiia is mainly a tropical African genus* 



The extreme limits of its 



distribution are from Lat. 9^ N, {B. Ileudelotiaim in Southern Senegambia) 

 to Lat. 30° S. {B. racemosa in Natal), and from Long. 13^ AV. (Senegambia) 

 to Long. 115^ E. [B. horneeusis in North Borneo). The species which lie 

 outside the main area arc B. racemosa in Natal (the only species outside the 

 tropics), B. capparuiifoUa in Northern Sladagascar^ and i>\ homeensis in 

 Borneo. The occurrence of the last at so great a distance from the continent 



