IN GYMNOSPERMOTJS PLANTS. 307 



of a scale, as in Fig. 391, or sometimes of a different shape (Fig. 



407), which bears two or more ovules upon some part of its mar- 



the axis, and not to the carpellary leaves ; and a one-celled ovary, with one 

 or more ovules arising from the base of the cell, would nearly represent the 

 typical state of the gynaecium. This theory, which the intelligent student 

 may easily apply in detail, offers the readiest explanation of free central pla- 

 centation, especially in such cases as Primula, &c., where not the slightest 

 trace of dissepiments is ever discoverable. It must be admitted that the mon- 

 strosities which occur in Primula, and some other plants with free central 

 placentation, favor this new view. It is also perfectly applicable to ordinary 

 central placentation; where we have only to suppose the cohesion of the in- 

 flexed margins of the carpellary leaves with a central prolongation of the axis 

 or receptacle which bears the placentae. But in case of parietal placentation, 



FIG. 394. Carpellary scale of Cupressus sempervirens (the true Cypress), seen from within, 

 and showing the numerous orthotropous ovules that stand on its base. 395. Branch of Abies 

 Canadensis (Hemlock Spruce), with lateral staminate flowers, and a fertile strobile. 306. Stam- 

 inate ament, magnified. 397. Carpellary scale of a fertile areent, with its bract. 398. Simi- 

 lar fertile scale, more magnified and seen from within ; showing the two ovules adherent to its 

 base: one of them (the left) laid open. 399. The scale in front, nearly of the natural size, its 

 inner surface occupied by the two seeds. 400. Polycotyledonous embryos of Abies and Cypress. 

 401. Vertical section of one. 402. Strobile of Taxodium distichum (Suborder Cupressineae). 



