202 MANUAL OF BOTANY 



in the centre and develop placentas from their margins in the 

 same manner as in ordinary axile placentation ; subsequently 

 the walls of the ovary grow more rapidly than the dissepiments, 

 so that the connection between them is soon destroyed ; and 

 from this cause, and also from the great subsequent development 

 of the placenta, the septa ultimately become almost or quite 

 broken up, so that the placenta is left free in the cavity of the 

 ovary. This view is strengthened by the fact, that in several 

 of the CaryophyllacesB we often find dissepiments in the young 

 ovary {fig. 435), and even traces of them at the lower part of the 

 mature ovary; these are the remains of dissepiments which 

 have become ruptured on account of the unequal development 

 of the parts of the ovary. The second type is illustrated by the 

 Primrose, and many other plants, which have a free central 

 placenta ; in these no traces of dissepiments can be found at any 

 period of the growth of the ovary. Duchartre and others who 

 have traced the development of the ovary in the Primulaoeae, 

 state that the placenta is free in the centre from its earliest 

 appearance, that it is originally a little papilla on the apex of 

 the thalamus, and that the walls of the future ovary grow up 

 perfectly free, and ultimately enclose it. The formation of such 

 a free central placenta cannot therefore be well explained upon 

 the marginal theory, as the carpels have never had any connec- 

 tion with it except at their bases. Nor is it necessary to suppose 

 that the placenta is essentially of foliar origin. Since it has been 

 established that the ovule is to be looked upon as the megaspo- 

 rangium of the flowering plant, it follows that it may be pro- 

 duced, as other sporangia are in some of the lower forms, either 

 Irom the axis itself, or from foliar outgrowths from it (sporo- 

 phylls). The case of the Yew, already referred to, is another 

 instance of axial origin, the ovule being here naked and terminal. 

 Other illustrations have been already given, and to them the 

 case of the Plumbaginacese may be added. InArmeria, a genus 

 of that natural order, the ovary, which is composed of five 

 carpels, surrounds a single ovule, which, rising from the axis, is 

 supported on a stalk which curves downwards at its apex, and 

 thus suspends the ovule free in the centre of the cavity {fig. 440). 

 Here, therefore, the ovule and placenta are developments of the 

 axis. 



EVom aU that has now been stated, we may draw the follow- 

 ing conclusions, namely : — that no single theory sufficiently 

 accounts for the production of the placenta in all cases ; but 

 that the axile, parietal, and some forms of the free centra] 



