OF ARTS AND SCIENCES. 319 



Notes vpon some Polynesian Plants of the Order Loganiacecs. 



By Asa Gray, M. D. 



These will naturally enoiigli follow the foregoing notes upon Ru- 



hiacece. 



1. Geniostomece (Geniostoma and Labor ded). 



Under the name of Hcemospermum, Reinwardt, Blume long ago cor- 

 rectly described the pulpy placentae of Geniostoma, which enclose the 

 seeds tind fill the cells of the fruit. It is remarkable that this should 

 have escaped the notice of succeeding authors, even of Alphonse De 

 Candolle, when he reduced Hcemospermum to a species of Genio- 

 stoma upon the authority of an authentic specimeij ; and of Bentham, 

 when he recently revised the Loganiacece* and had Blume's detailed 

 character of Geniostoma before him ; t especially as in dried speci- 

 mens the seeds all cohere in a mass, and are covered with a pellicle 

 of dried pulp, which, when soaked, promptly swells up and encloses the 

 seeds, just as they doubtless are enclosed in the fresh fruit. It appears 

 from a supplementary note to Mr. Bentham's paper, that this struc- 

 ture has been noticed by M. Bureau, in his inaugural dissertation upon 

 this group of plants (which I have not seen) ; but I do not well under- 

 stand what is meant by " the curious expansions of the placenta " in a 

 " regular stellately-Iobed form." I am confident that there are no 

 distinct arilli severally enclosing the seeds, in the manner of Podo- 

 phyllum, but an equable pulpy development of the placenta. In this 

 lifht it is questionable if Bentham's G.fagrceoides (the flowers of which 

 are unknown) is really a Geniostoma. Although the pericarp is cap- 

 sular, the analogy of Geniostoma to Gardenia and the Randiece in the 

 placentation is not to be overlooked, especially as the a3stivation of the 

 corolla is similar. To this there are exceptions, however. Good as 

 characters of osstivation are in Ruhiacece, and as they are likely to be 

 in these " Ruhiacece with a free ovary," we can never trust implicitly to 

 the difference between the imbricative and the contorted or convolute. 

 No one can be more aware of this than Mr. Bentham, who will not 

 be surprised to learn that this distinction between the second and third 

 sections of his tribe Eidoganiece is untenable, — Geniostoma rupestre 

 having occasionally the {Estivation assigned to Logania, namely, with 



* In Journal of the Proceedings of the LinnsEan Society, Vol. I. 1857. 

 t In Mus. Bot. Lugd.-Bat., Vol. 1, p. 238, 1850. 



