232 MORPHOLOGY OF THE ANGIOSPERMS 



column of tissue between the carpels is interpreted as receptacular. The 

 "receptacle" in this description is actually the fused margins of the 

 carpels, as demonstrated by anatomy. 



Septal Nectaries. Where lateral union between carpels is incomplete, 

 the unfused areas commonly become secretory, and well-defined nec- 

 taries are formed. These nectaries are common in the Liliaceae and 

 related families and occur occasionally in the palms and other genera. 

 The secretions of these glands may exude along longitudinal slits be- 

 tween carpels or at tlie top of the ovary when the carpels are fused 

 along the dorsal margin. These nectaries may have elaborate form and 

 exude the secretion through a specialized canal leading to a small open- 

 ing on the top or side of tlie ovary — Cocos, Ananas. The septal nectary 

 seems to be restiicted to monocotyledons and has probably arisen in- 

 dependently in members of this taxon. 



The presence and position of the unfused areas of carpel walls in the 

 syncarpous gynoecium is related to the history of carpel connation. The 

 transformation of the semienclosed pockets between the carpels into 

 nectaries and the elaboration of their form constitute an excellent ex- 

 ample of adaptation of flower structure to insect pollination. 



Coenocarpous and Paracarpous Gynoecia. Apocarpous and syncarpous 

 are terms long in common use to describe gynoecia consisting of free 

 and of fused carpels, respectively. But attempts have been made to 

 supplant "syncarpous" by introducing a new term, coenocarpous, re- 

 viving an old term, paracarpous, and redefining syncarpous. In the new 

 treatment, gynoecia made up of united carpels are coenocarpous; those 

 with a single ovarian chamber, with placentation parietal, free central, 

 or basal, are paracarpous; only those with two or more separate cham- 

 bers and longitudinal fusion incomplete are syncarpous. Syncarpous 

 gynoecia, under this definition, may show three structural zones: a 

 fertile base, the ovary, which is syncarpous; a median part, the style, 

 which is paracarpous; and a stigma, apocarpous. These terms, so used, 

 are not good, because coenocarpous and syncarpous have much the 

 same meaning, and the classification is valueless, because "paracarpous" 

 and "syncarpous" forms are considered morphologically distinct types. 

 No line can be drawn between paracarpy and syncarpy; in some 

 families, all transitional stages occur, and, in some taxa, the ovary is 

 "syncarpous" in the basal part and "paracarpous" in the upper part. 

 Yet the new terms have been accepted and used in publications in 

 continental Europe. 



The term false coenocarpy has been proposed to describe gynoecia 

 where adnation of carpels to a cup-shaped receptacle ties together 

 carpels otherwise free — Hydrocharitaceae, Pomoideae, Biitomus. The 

 term seems to be applied also to gynoecia where the carpels are held 



