106 Linnean Society. [Dec. 17? 



other so as to become opposed to the persistent posterior carpellum ; 

 in Agrimonia and Spiraa (when dicarpous) from a similar . sup- 

 pression ; as also in reduced fruits of Reseda luteola, &c. ; 3rdly, ob- 

 lique, which he describes as of frequent occurrence both in plants in 

 which the carpella are generally anterior and posterior, and in those 

 in which they are as predominantly right and left, and which he sup- 

 poses to arise from the remaining lateral carpellum of a tricarpous 

 ovary retaining nearly its original position when the other lateral 

 carpellum has disappeared, in consequence of which the posterior 

 carpellum is somewhat displaced, becoming obliquely posterior. He 

 regards the single carpellum as the result of the non- development 

 of one of the carpella of a dicarpous ovarium, and its position may 

 consequently vary in three different ways : 1st, anterior, as occurs 

 in 1-carpellary ovaries oi Myrtacece., Onagrariee, Polygalea, Legumi- 

 nos(B and Acanthaceee, to which may probably be added Hippwidece, 

 Bruniacece, &c. ; 2ndly, j^o^^mor, as in the 1-carpellary ovaries of 

 Houttuynia cordata and Piperacece ; 3rdly, lateral or oblique, instances 

 of which occur in Morea, in Elatostemma, and in Celtidece. The 

 normal number of carpella in all ovaries he regards as three or a 

 multiple of three ; the additional series being frequently reduced 

 by abortion in the same manner as the first, and thus giving rise to 

 the formation of ovaries with four and five carpella. Tricarpous 

 ovaries generally have their component parts placed two laterally 

 and one posteriorly ; but exceptions to this rule occur, as for exam- 

 ple in Viola, where the third carpellum is anterior, and in Clethra, 

 Pittosporum and Delphinium, in which the position of the carpella 

 varies in the same plant. 



Mr. Clarke next proceeds to consider the value of the characters 

 derived from the position of the carpella, for which purpose he has 

 framed a large table containing the results of long-continued ob- 

 servations on a multitude of exogenous plants with monocarpous or 

 dicarpous ovaries. In this table he constitutes two primary divisions, 

 viz. Proterocatyous, in which the carpella when single are anterior 

 or lateral, never posterior ; and Heterocarpous, in which the single 

 carjoellum is for the most part a mixture of lateral, anterior and 

 posterior, and is rarely wholly posterior. The position of the com- 

 ponent parts of the dicarpous ovarium also appears to be more per- 

 manent in the first than in the second division. From this table 

 Mr. Clarke deduces various inferences in relation to the systematic 

 arrangement of plants, and the importance of the characters derived 

 from the position of the carpella, and more especially from that of 

 the single cai'pcllum, which is liable to fewer and less important 



