68 



OEGANOGRAPHY AND GLOSSOLOGY. 



396. NigeUa. 



Ripe OTary 



cut transversely. 



398. Lychnis. 



Young ovary (mag.) 



cut transversely^ 



period is divided into superimposed cells by septa formed of the parenchyma of the 

 ovary, which is intruded horizontally between the seeds. 



Spurious cells {loculi spurii) are cavities in the ovary which do not contain seeds. 

 The young ovary of NigeUa presents five cells, each containing two piles of ovules ; 

 later (fig. 396) there appear ten cells, of which five in the centre of the fruit contain 

 seeds attached to their interior angle ; the other five are exterior 

 to these, and are empty, and due to the inflation of the epicarp 

 (bp), which in swelling has dragged with it the mesocarp (m), whilst 



the endocarp (en) has re- 

 mained in its place. 



Central placentas are 

 said to be free {p. centrales, 

 liherce), when they are not 

 united by septa to the walls 

 of the ovary, and appear com- 

 pletely independent of the 

 carpels ; this placentation is 

 characteristic of Primulaceoe [Pimpernel, Primrose, Cyclamen, fig. 397). To explain 

 this isolation of the placentas, it is assumed that the edges of the earpellary leaves 

 join throughout their length, and constitute a one-celled ovary, but that their basal 

 edges dilate, and ascend in the middle of the cell to form a central mass of 

 placentas. The placentas of Primulacece are thus confined to the bases of the 

 carpels. The reverse is the case in the one-celled ovary of Combretacew, where the 

 ovules spring from the top of the cell. 



In most Caryophyllew (Pinh, Lychnis), the placentas appear to be free, but this 

 arises from the early evanescence of the septa, which can only be well seen in the 

 very young flower (fig. 398). 



Some German and French botanists regard the earpellary leaf as a pi'oteetive 

 organ merely ; denying that it has the power of producing buds, and limiting this 

 power to the floral axis. According to these, the axis alone produces Ovules, and 

 the earpellary leaves protect them. In the case of many-celled ovaries, they regard 

 the edges of the earpellary leaves as folded inwards till they reach and cohere 

 with the axile placentas (which in no wise belong to them), the fibro-vascular bundles 

 of the placentas losing themselves in the tissue of the styles, which are continuations 

 of the midribs of the carpels. In unilocular compound ovaries they consider that the 

 placentiferous axis branches like the spokes of a half-opened parasol, and that the 

 branches run along the contiguous edges of the earpellary leaves (Heartsease, Mignon- 

 ette, fig. 384 ; Orchis, fig. 385). 



This modification of the earpellary theory of placentation rests on the isola- 

 tion of the placentas in Primulacece (fig. 397) ; on the enormous disproportion 

 of the placentas relatively to the cai'pellary leaves in various plants [Lychnis, 

 fig. 398 ; Canipanula, fig. 390) ; and on the arrangement of the nerves in certain 

 ovaries [Pea, fig. 14 ; Columbine, fig. 13), wherein two systems of fibro-vascular 

 bundles are distinctly visible ; the one coming from the median nerve, the others 



