194 Mr. B. Clarke on Relative Position ; 



frequently lateral. The position of the fertile cell therefore (in 

 such cases) is subsequently noticed only to show how far the 

 inquiry has extended. 



But there is a circumstance occurring not unfrequently in 

 compound ovaries, by which the position the single carpel would 

 occupy may be with confidence assumed, viz. the diminished size 

 of one of the carpels. Thus, in Circcea alpina, when the ovary is 

 two-celled the posterior cell is both shorter and less in its dia- 

 meter, and when one-celled the cell is always anterior, and ana- 

 logous examples occur in Stylidium and Dumpier a. In Vale- 

 rianacefe the barren cells are sometimes so reduced as to be 

 scarcely apparent, but occasionally they are larger than the fer- 

 tile ; and although this is an exception as to the size of the carpel, 

 yet like the smaller cells they are destitute of ovules. 



One of the stigmas of a dicarpous ovary being larger than the 

 other is a character which is likely to prove of the same value, 

 as in Labiatse and Verbenacese the anterior portion of the stigma 

 is sometimes enlarged, and in Lantana and Lippia the two-celled 

 ovary is formed by a single carpel anterior. In Acanthacese, also, 

 stigmas occur with the anterior lobe elongated, and in the one- 

 seeded Mendozia it is the anterior carpel which is fertile*. 



To the inequality of the stigmas there are however exceptions 

 (which may be compared to the barren cells of Valerianella 

 having become inflated, or to sterile stamens having become 

 petaloid), as in Leutibularieae the posterior portion of the stigma 

 is constantly larger, and also in Polygala speciosa. But as it is 

 always the posterior lobe which becomes enlarged, while in Poly- 

 galacese it is the posterior cell which is suppressed in one-celled 

 ovaries, these two exceptions are unimportant ; and the larger 

 lobe of the stigma being variable in its position should alone 

 perhaps be taken as an evidence of variation in the position of 

 the single carpel. Schweiggeria also supplies another instance 

 (though less marked), in which the two larger stigmas are late- 

 ral or obliquely posterior, while the larger-ribbed carpel of the 

 tricarpous ovary is anterior. 



Fkee Central Placent.b. 



It would not perhaps be expected, that in compound ovaries 

 having a free central placenta, the position of the ovule when 

 solitary would supply any evidence from which the position the 



* In M. puherula the fertile carpel is always anterior, and from the simi- 

 larity of the species there seems no reason to doubt but that it is so in all. 

 The two-celled fruit described by Martius as having its cells placed one 

 above the other, is produced by an extension of the placenta across the an- 

 terior cell, as the remains of the posterior cell are behind the upper fertile 

 cell. 



