TRELEASE N. AM. RHAMNACEiE. 359 



inserted, and (excepting some Celastracea?) one or two basi-fixed 

 ovules with a ventral raphe, for each carpel. The Rhamnaceoe 

 difler from both of the other Orders in their valvate calyx, the lobes 

 of which are mostly keeled or crested within, and in having their 

 stamens alternate with these, hence opposite (and usually clasped 

 by) the petals when they are present. From most of our Celas- 

 traceffi they also differ in the uniform absence of an aril ; and 

 from the Ampelidete, in their drupaceous or dehiscent fruit, rela- 

 tively larger embryo, and in never producing compound leaves, 

 nor tendrils opposite the leaves. 



In most species of Rhamnaceaj the flowers are hermaphrodite. 

 Exceptions are found in Gouania and one section of R/iajiunis^ 

 where, by suppression of stamens or pistil, they become essen- 

 tially dicEcious as in Hiioiiymits^ Ilex, Acer, etc. Mr. Darwin 

 has called attention to the curious differences in length of the pis- 

 til in the dicEcious Rka?i/nits cathartica, which bears long- and 

 short-styled staminate, and long- and short-styled pistillate flow- 

 ers. This unexplained differentiation apparently dates back to a 

 time anterior to the separation of the sexes by abortion of one or 

 the other of them. Long- and short-styled flowers of R. lanceo- 

 lata — first carefully observed, I believe, by Professor Porter — 

 are described by Dr. Gray.* Our species of RJianimis of the 

 hermaphrodite section Frangula, are likely to prove protan- 

 drous, like the European R. Frangtila\ and Paliitriis aculea- 

 his. The variability of R. crocea in the presence or absence 

 of petals, is to be compared with that noted by Mueller for R. 

 pumila^X though it may not be explicable on the same grounds. 

 So far as our flora is concerned, the entire Order seems to be 

 adapted to pollination by small Hymenoptera, Coleoptera, and 

 Diptera, but observations on this point are wanting. 



A climbing habit has been acquired in Bercheinia and Gou- 

 ania, the former twining, the latter having twig-tendrils. 



Dissemination by birds is apparently provided for in those 



* Darwin: Different Forms of Flowers. — Index, 

 \ Mueller: Befruchtung; der Blumen, 152. 

 \ Alpenblumen, 169. 



