MR. BETfTlIAM'S NOTES ON HOMALIUM., 31 



Notes ou Ilomalium. By Gteorge Bentham, Esq., V.P.L.S. 

 [Read June 2nd, 1859.] 

 In the revision of the 'Hong Kong Elora,' with which I have been 

 lately occupied, I had occasion to investigate the validity of the 

 genus Blachicellia, of which an elegant Chinese species was many 

 years since introduced into our gardens, and has found its way 

 into catalogues and botanical works under six different names. 

 The result of this investigation has been the conviction that the 

 genus must altogether be united with Homalitim. I could not, 

 however, come to this conclusion without a careful examination of 

 all the species referred to both genera of which we had specimens ; 

 and as I also found several unpublished ones in the herbaria at 

 Kew, I have been led to draw up a short synopsis of the whole 

 group, which I now beg to lay before the Society. 



The genus Blachwellia was originally established by Jussieu 

 and Lamarck on two or three Mauritius trees which differed from 

 the only two Homalia then known, both from South America, in 

 having only one instead of three stamens opposite each petal (or 

 inner segment of the perianth as it was then called). Ventenat 

 afterwards added two or three eastern species having the same 

 peculiarity ; and De Candolle, in the 2nd vol. of the ' Prodromus,' 

 maintained the two genera, with the same technical character as 

 well as geographical distinction. Since then, several Asiatic 

 species with two or more stamens to each petal have been de- 

 scribed, and yet they have been published as Blachwellias. An 

 exception, has been made in the case of the African Homalitim 

 angustifolium, which has the character of the American ones; 

 and Sir James Smith, who always closely adhered to generic 

 character, published it accordingly as a Homalium. Modern 

 botanists, however, applying too literally the rule of " character 

 non facit genus," appear in this instance to have practically 

 adopted geographical origin as the sole basis of the limitation 

 of the two genera. Some other characters are indeed indicated 

 by Endlicher and others, such as the supposed larger calycine 

 segments or capitate stigmas of the American ones ; but none 

 of them will bear the test of examination. And few, I believe, 

 would now contest the generally admitted rule in systematic 

 botany, that geographical origin without any character is not 

 to be recognized as a generic distinction. It therefore becomes 

 necessary to unite the Asiatic clustered-stamened species with 

 the corresponding American Ilomaliums. As some of these, again, 

 in other respects resemble the single-stamened species more than 



