644 DR DICKSON ON DEVELOPMENT OF 



the ovarian wall is completed posteriorly, by the coalescence of the extremities 

 of the original semilunar elevation, the organic centre of the receptacle is some- 

 what depressed. Almost as soon, however, as the ovarian wall is complete, the 

 receptacular centre enclosed by it begins to be developed as a more or less hemi- 

 spherical protuberance — the young placenta. At no period of its development has 

 it any connection with the ovarian wall : it is as truly " free-central" as that in 

 Primulacese. The ovules make their appearance first on the top of this hemi- 

 spherical placenta, and continue to appear in succession from above downwards, 

 until the surface is covered by them (Plate XXIX. fig. 16). This placenta does not 

 exhibit the slightest trace of the barren apex, which is so characteristically present 

 in that of Primulacese — not even a bare spot,— but is uniformly and densely 

 crowded with ovules over its whole surface. The ovules originate as small 

 mammillae, which become invested with a single integument, and undergo the 

 anatropal curvature, as represented in the series given in Plate XXIX. figs. 17-22. 

 They are placed so that the raphe is superior where the ovules project horizon- 

 tally, internal where they have an upward direction, and external where they 

 have a downward one. 



A hnormalities. 



In the course of the examination of numerous flowers, for the purposes of the 

 foregoing investigation, I have met with a considerable number of cases of abnor- 

 mality or monstrosity, some of which I think worthy of being recorded. 



In Plate XXX. figs. 23 and 24, are represented two cases of remarkable modi- 

 fication in the symmetry. In fig. 23, the flower is dimerous and regular, with two 

 sepals, two petals, and two stamens, in decussate succession. The ovary here is 

 as yet only faintly indicated * In fig. 24, there are six sepals, of which one is 

 anterior, one is posterior, and four are lateral, these last being conveniently 

 distinguishable as antero-lateral and postero-lateral. Alternating with the sepals 

 are six petals. There are five parts of the andrcecium, viz., two fertile stamens 

 superposed to the antero-lateral sepals, and three staminodes, of which two are 

 superposed to the postero-lateral sepals, and the third is placed between the two 

 fertile stamens, and thus superposed to the anterior sepal. 



The other abnormalities I have figured are some very interesting ones affect- 

 ing the pistil. In fig. 25, the posterior wall of the ovary is deficient, the 

 placenta and ovules being exposed ; the result, doubtless, of imperfect coal- 

 escence of the extremities of the primitive semilunar elevation, a defect of 

 development analogous to spina bifida, cleft-palate, hypospadias, &c, in the 

 animal subject. In fig. 26, the posterior (small) lip of the stigma is seen 

 to be bipartite. In this, as in the last abnormality, we have impressed upon us 



* This flower was unfortunately detached before I had ascertained whether the sepals weie 

 anteroposterior or lateral. 



