206 



is liable to fewer and less important exceptions. Thus for instance 

 he considers the posterior position of the single carpellum of Ceraio- 

 phylleae, corresponding as it does with that of Piperaceae and their 

 allies, and differing as far as known from that of any other order with 

 which it could be associated, as a strong argument of affinity. He 

 refers to the case of two-celled ovaries with unequal cells, and regards 

 the sujterior development of the larger cell or of the corresponding 

 stigma as indicative of what would be the position of the single car- 

 pellum, were the ovary to be so reduced. These remarks are followed 

 by observations on the general character of his divisions and subdivi- 

 sions, and by some notes on the position of carpella as regards endo- 

 genous plants and Rhizanthese, and on the relation of didynamous 

 stamens and carpella as regards their order of suppression ; and the 

 first part of the memoir concludes with some remarks on the difficulty 

 of determining with precision the true axis of the' inflorescence, and 

 the means of obviating this difficulty in certain cases. 



The second part of the memoir is more especially devoted to the 

 consideration of ovaries consisting of a single carpellum, to the rela- 

 tions borne by this carpellum to the axis in various families referred 

 by the author to each of his two principal divisions, and to the 

 grounds from which this relation is deduced. This being entirely 

 matter of detail is scarcely susceptible of analysis, but some of the 

 incidental observations connected with it may properly be noticed 

 here. Mr. Clarke states that in Scleranthus annuus the funiculus is 

 uniformly posterior to the seed and on the same side with the coty- 

 ledons, in which character that plant differs from Chenopodeae and 

 Amaranthaceae, and as far as he has been able to ascertain from Ille- 

 cebreae, in which the funiculus is either anterior or lateral, and the 

 cotyledons (in pendulous seeds) on the opposite side of the seed or 

 less frequently lateral. Of thirty-two ovaries of Circaja alpina, thir- 

 teen had two cells with an ovule in each, but the posterior cell con- 

 stantly smaller than the anterior, in twelve the posterior cell was 

 empty, and in seven entirely absent; and this analogy with some par- 

 ticularities in structure led him to regard the single cell of Hippuris 

 as most probably resulting from a single anterior carpellum. He 

 shows by a series of diagrams that the position of the fertile cell in 

 Valerianeae is always lateral and external; and observes that in the 

 genera with an irregular corolla it always bears the same relation to 

 the irregularity of the flower. He infers from an inferiority of deve- 

 lopment of the posterior carpellum in Stylidium graminifblium, that if 

 the ovary in that genus were reduced to a single carpellum, that car- 



