364 MOKPHOLOGY. 



c. The accessory foliar Organs of the Flower. 



156. Besides those parts of the flower of which we have 

 already spoken, other foliar organs are frequently met with, 

 which, considering their simple structure (scales of varying degrees 

 of thickness) or very aberrant shape, may be regarded as parts of 

 the flower but partially developed. According to their position, 

 I distinguish two forms : 1st, from the outermost floral envelopes 

 to the outermost circle, exclusively, in which stamens are deve- 

 loped, the accessory corolla or paracorolla, and its accessory petals 

 or parapetala; and, 2dly, from that circle, inclusively, to the gernien, 

 the accessory stamens, or parastemones. 



The accessory corolla consists sometimes of scales, which are 

 either thin and leaflike or thick and fleshy, with their margins 

 sometimes entire, sometimes divided, as in the Grasses, the inner 

 tri-merous circle of foliar organs, of which one member is com- 

 monly abortive : in Vallisneria the three little scales. More fre- 

 quently the accessory corolla exhibits very peculiar aberrant forms, 

 which may present the forms of the floral envelopes in miniature, 

 and often distorted, as in the long, thin foliaceous organs in the 

 flower of Aconitum, which imitate a long-clawed, spurred peri- 

 anthial leaf, the horn-shaped accessory petals of Helleborus, Trol- 

 lius, Nigella, &c., the extraordinary little, mostly boat-shaped 

 leaflets of the Loasacecs. I am unacquainted with any instance in 

 which the members of the accessory corolla are coherent. The 

 structure is either very simple, as in the generality of scales, which 

 consist merely of delicate cellular tissue ; or it resembles that of the 

 floral envelopes and its appendages. The secretion of nectar at 

 given places is very frequent here, especially in concave forms. 



Accessory stamens occur in two ways either as perfectly free 

 foliar organs, or wholly coherent. a. In the first case they 

 are more or less similar in form to the stamens, and often, espe- 

 cially if (as in Chelone and Scrophularia) they belong to a circle, 

 of which some members are developed to stamens, they are formed 

 exactly like a filament without its anther, as in many Geraniacece : 

 sometimes they are also scale-like here, as in Veronica, where they 

 represent two parts of the four-membered staminal circle.* When 

 they form a proper circle by themselves, they are usually developed 

 as small scales, as in Pimelea, Gnidia, &c. 



b. In the second case, they mostly constitute the so-called in- 

 ferior ring (annulus hypoyynus\ and are then usually thick, fleshy, 

 and juicy, as in Daphne, Celosia^[, and Trapa. Sometimes this 

 ring is lobed, so as to exhibit the number of its members clearly, 

 as in the generality of the Ericaceae : in Chrysosplenium it is eight- 

 lobed ; in Cobcea scandens and Convolvulus five-lobed ; or indis- 



* This condition also appears to exist in Lathraa and Orobanche. 

 f In which this ring has hitherto been wholly overlooked. 



