mh. g. bektiiam ok myrocabpus. 265 



The observations of Sen. Correa deMello would perhaps have 

 determined the removal of Myrocarptis to the Sclerolobiese, if the 

 aestivation described by him were constant. With the above Notes 

 he remitted several racemes in bud, which have been carefully ex- 

 amined. In them we found the estivation of the petals still more 

 variable than is mentioned by Mello ; for the upper, or posterior 

 petal, although occasionally inside, as he describes it, is more fre- 

 quently partially outside, overlapping one of the lateral petals on 

 one side, and being overlapped by another lateral petal on the 

 other side. In one flower, at least, the upper petal was entirely 

 outside ; and in another the estivation was contorted, each petal 

 overlapping an adjoining one on one side, the imbrication in all 

 cases slight, and only observable in the upper portion of the bud, 

 Myrocarpus^ then, like one species of Stveetia where a similar di- 

 versity of estivation has been observed, must be regarded as oue 

 of the few connecting links between the large suborders Papilio- 

 naceae and Csesalpiniese ; and as in other respects it is evidently 

 more closely allied to Myroxylon than to any Sclerolobieae, we 

 should prefer leaving it in the place assigned to it in our ^ Genera 

 Plantanun,' between Myroxylon and Sweetia. 



The course of expansion of the flowers, from the apex to the 

 base of the raceme, pointed out by Correa de Mello is unusual in 

 Leguminosse, but not quite exceptional, nor yet does it appear to 

 be of any systematic importance ; for it has been observed, for in- 

 stance, in two or three Australian species of Crotalaria^ whilst the 

 development is normal in closely allied species. 



Since writing the above, I have received M. Baillon's obser* 

 vations on Csesalpiniese, in the first vol. of his ' Histoire dee 

 Plantes,' from which it appears that he has found in Cadia, 

 Porsk., and in Barklya, F. Muell., the same variable aestivation 

 as that observed, as above, in Myrocarpus. He therefore proposes 

 to remove those two genera to CaesalpinieaB, although they have 

 both, in a very decided manner, the hooked radicle of Papilio- 

 nacese. It still appears to me, however, a more natural arrange- 

 ment to maintain the Sophorese as limited in our ^ Genera Plan- 

 tarum/ as a somewhat variable group, including a large pro- 

 portion of very distinct monotypic, or almost raonotypic, genera 

 probably of great antiquity and constituting in various ways con- 

 necting links between' the two great suborders Papilionacea and 

 G<B8alpini€<B, but yet more nearly allied to the former than to the 

 latter. 



G. Bentham. 



