Bibliographical Notices. 215 



dentally observes that the two genera, Hernandia and Inocarpus, of 

 which that family has been composed, do not appear to be very 

 nearly related to each other. 



The fourth species figured in this part is Cardiopteris lobata. 

 Wall. List, which is identified with Cardiopteris Javanica of Blume. 

 After tracing some curious points in the botanical history of the ge- 

 nus, Mr. Brown proceeds to notice the more remarkable peculiarities 

 of its structure, and discusses the questions of its hermaphroditism, 

 the position of the micropyle of its seed, and the singular arrange- 

 ment of its perfect and imperfect stigmata. In illustration of the 

 latter point we quote his description of the pistillum. " The external 

 structure of the pistillum is very singular. In an early stage of the 

 flower, immediately before or even at the time of expansion, there 

 are apparently two stigmata : of these the more obvious is capitate, 

 undivided, fleshy, but not papillose, and is supported on a distinct 

 style ; the second is quite sessile, much shorter in this stage than the 

 capitate branch, and having its upper or inner surface distinctly stig- 

 matic or papillose. In the next stage, the latter, which I regard as 

 the eflicient stigma, gradually enlarges, becoming longer than the 

 capitate organ, which in my opinion is an imperfect stigma, and as in 

 this stage the ovarium though enlarged has not perceptibly increased 

 in diameter, this capitate stigma has the appearance of being lateral. 

 The perfect stigma, which continues to lengthen, its upper surface 

 becoming more evidently hispid or papillose, not unfrequently re- 

 mains crowning the samara even when ripe, but frequently also it 

 is then deciduous, while the imperfect capitate stigma, which has 

 undergone no change either in size or surface, more generally remains 

 after the real stigma has fallen." With respect to the affinities of 

 this curious genus, Mr. Brown does not regard any of the approxi- 

 mations hitherto made as satisfactory ; and although aware of several 

 important objections to the view, is " inclined to consider Cardio- 

 pteris as an isolated genus or family to be placed at no great di- 

 stance from Phi/tocrenece, chiefly through lodes." 



The concluding article contains a monograph of the genus Ben- 

 nettia, established by Mr. Brown in Dr. Wallich's List in 1847, and 

 recently published by M. Tulasne under the name of Cremostachys. 

 Mr. Brown regards Bennettia as bearing " the same relation to Anti- 

 desma (for Antidesmece contains at present no other well-established 

 genus) that the polypetalous bear to the apetalous genera of Euphor- 

 biacece." In the present case he states that " the presence of petals 

 may even be regarded as of more than ordinary importance, their 

 usual form in the male flower being necessarily connected with the 

 aestivation of stamina." This remarkable peculiarity is thus desciibed 

 in the species figured, Bennettia Javanica, R. Br. : — " Stamina decern 

 distincta. Filamenta brevissima, latiuscula, sepalis et petalis oppo- 

 sita. Antherae biloculares, loculis connectivo lato distinctis longitu- 

 dinaliter dehiscentibus ; omnes cucullis petalorum semi-inclusse, ita 

 ut duse petalo singulo oppositse esse videantur (exterior interiorque), 

 sed dum exterior ad filamentum petalo oppositum pertinet, interior 

 e loculis respondentibus filamentorum duorum petalis alteniantium 

 formata est." Mr. Brown adds, that " the affinity between Euphor- 



