THE GENUS AMARALIA. 



Amaealioid species or IIaxdta. 



In consideration of the confusion already referred to between 

 certain species of Bandia on the one hand, and the genus Amaralia 

 on the otlier, it is desirable to deal with these species in this place. 



M. Hua makes frequent reference in his paper on Sherbouruia, 

 already cited (supra, p. 1). to a Bandia amaralloides K. Schum,, 

 which had appeared previously only in manuscript, as the determina- 

 tion on the labels of several of Zenker's Camc;roon plants ; among these 

 w^ere more than one species of Amaralia. M. Hua quotes, as a type 

 of this, Zenker 1017 ; this, however, is undoubtedly identical \\i\\\ 

 Staudt 12 ! — also a Cameroon plant — the type of B. streptocaulou 

 K. Schum. in Engler, Bot. Jahrb. xxiii. (1897) 440. This has an 

 obscurely shortly - toothed calyx - limb ; in some other respects it 

 resembles an Amaralia — particularly in the climbing habit, the 

 stipules, the shape and venation of the leaves, the shape and 

 indumentum of the corolla, and the shaj^e and surface of the fruit. 

 The resemblance, in fact, is so close as to tempt M. Hua {loc. cif.) 

 to include this B, amaralioides in his SJierbournia ; he does not, 

 however, quite succumb to this temptation. It seems clear, I think, 

 that this resemblance is due to homoplas}', to similarity of form 

 induced by similaritj^ of habit and habitat ; most species of Amaralia, 

 Ave have seen, are scandent shrubs, trie climbing being helped largely 

 by the strong petioles ; and the habit of B. streptocaulou and its im- 

 mediate allies is precise!}^ sunilar. The toothing of the calyx, ceteris 

 paribus, would not presumabl}^ be affected in descent, at any rate so 

 rapidly as the vegetative parts, l)y this habit ; wherefore its diiterences 

 in these cases provide a critical characteristic, determining a generic 

 race descended from a Bandia-\\ke ancestor from which arose also, 

 later in evolutionary history, B. streptocaulon and its near relatives. 



This conclusion receives no little support from the existence 

 of other " Amaralioid " species ; three have been described before, 

 thi-ee others are, I believe, new. For the sake of completeness 

 I include all the AmaraKoid species, old and new, in the following 

 systematic account, preceded by a clavis. All these seven Baudia- 

 species have a persistent tubular cah'x-limb with short, or in some 

 cases obscure, teeth. All the specimens in the National and Kew 

 herbaria are quoted below. 



These species display an interesting parallel series to those of 

 Amaralia. The relatively small-leaved B. streptocaulon corresponds 

 Avith A. Slierb our nice, in distribution as in other points, for the 

 Cameroons is its eastern and southern limit. B. amaraliocarpa, 

 AN-ith prominently ribbed fruits, is paralleled by A. Zenhri, witli 

 similar distribution — /. e. Upper Guinea. B. curvipes goes hand-in- 

 hand with A. hignoniceflora in Angola. B. castaneofulva in Angohi 

 and B. hapalophylla and B. annulata in Upper Guinea are sharply 

 distinct in regard to their leaf -indumentum ; otlierwise they are veiy 

 Amaralia-\\\%. Lastly, B. cladantha, a tree-species of Nigeria and 

 the Cameroons, closely resembles the East-African A. peadulijlora 

 in habit, in vegetative system generally, and in the caulilloroi.s 

 inflorescence. 



