125 



XXV.-A REVISION OF DUBOUZETIA. 



T. A. Sprague. 



A sketch of the history of Duhouzetia was given in Kew 

 Bull., 1907, No. 1, pp. 10-11, where the genus was compared with 

 Tricuspidaria, and the arguments were stated for and against its 

 being generically distinct from the latter. The results of the 

 comparison were somewhat negative, owing to the lack of material 

 preventing the verification of certain alleged generic characters. 

 The matter remaining doubtful, the genera were kept separate 

 provisionally, and two new species of Duhouzetia were described 



in No. 2, pp. 57-58. 



Through the kindness of Dr. 0. Lignier, Professor at the Faculty 

 of Sciences, and Director of the Botanical Institute at Caen, we 

 have now been able to examine the material of Diibouzetia con- 

 tained in Vieillard's own herbarium, with the following results : 

 1, the confirmation of three generic characters, making it evident 



that Duhouzetia is distinct from Tricuspidaria ; 2, the reduction 

 of D. parviflora, Brongn. et Gris, to D. elegans, Brongn. et Gris ; 

 3, the description of an additional new species, of which imperfect 

 material already existed in the Kew Herbarium. 



At p. 10, it was stated the " capsule of D. elegans 



is loculicidal, with only a very faint indication of a septicidal 

 split, or none at all, so that the distinction drawn between Duhou- 

 zetia and Tricuspidaria, as regards the dehiscence of the fruit, 

 now breaks down." 



It now appears, however, in the light of the fresh material, that 

 the condition just described for D. elegans is merely the first 

 stage of dehiscence : the loculicidal split extends only a short 

 distance (the uppermost \-\ of the capsule) and stops ; and the 

 capsule then undergoes almost complete septicidal dehiscence, the 

 old fruit finally separating into open cocci, which fall off from a 

 persistent central column about the same time as the expansion of 

 the next set of flowers. Ry this time the pericarp has separated 

 into a brittle exocarp, and a tough, almost woody, endocarp 

 [ Vieillard, 2355, in Herb. Inst. Bot, Caen]. In D. feionema the 

 dehiscence of the capsule is apparently similar. 



In the figure of D. campanulata given by Brongniart and Gris 

 in Nouv. Arch. Mus. Hist. Nat. Paris, vol. iv. t. 13, a distinct 

 loculicidal split is shown, which extends from the apex about a 

 third of the way down the capsule, and their description of the 



dehiscence is as follows : — " Fructus septicide 



dehiscens, carpellis simul apice secundum nervum medium 

 incomplete fissis et marginibus placentiferis disjunctis, demum 

 omnino liberis et endocarpio lignoso incurvato ab epicarpio 

 secedente, eolumna centrali haud placentifera sola persistente." 



The dehiscence of the capsule of Duhouzetia, therefore, is 

 mixed, with predominance of the septicidal type ; that of Tri- 

 cuspidaria on the other hand, is purely loculicidal. 



The second generic character confirmed for Duhouzetia is the- 

 orientation of the carpels when all five of them are present, 

 namely, that they are opposite the petals, as stated by Szyszy- 

 lowicz; this has now been verified for D. acuminata ami 



D. leionema, n. sp. 



27805 



B 



