A PROBABLE ORIGIN OF EXISTING FLOWERS. 



43 



up the two margins in the pod ((not in ordinary leaves) give off branchlets 

 to the ovules. These two marginal cords and the midrib diverge from 

 a common vascular bundle in the stalk of the carpel. 



In the Gymnosperms, mentioned, the position of the ovules is not 

 upon the margins of the " carpellary scale," for they stand in front of 

 the scale. In Cycas, however, we find something much more nearly 

 like that of an Angiosperm ; for although the carpellary leaf is still 

 leaf-like, being pinnately divided, the ovules are borne along the two 

 margins ; but the blade is, of course, not closed over them. 



The nearest approach to a regular flower may be represented by 

 that of the fossil member of the Bennettites (the group approximating 

 the living Cycads) of the Secondary Epoch. Cycadeoidea (the name 

 implying "resembling a Cycas") has a whorl of hairy, lanceolate 

 bracts, surrounding a whorl of stamens, consisting of a main axis 

 with lateral anther-bearing pinnae, for it is apparently formed out of a 

 pinnate leaf. 



In the middle of the flower is a conical receptacle covered with 

 naked ovules. 



Leaving undetermined as to how ovules become formed within a 

 closed ovary for the future to discover, we will consider how the three 

 outer whorls of ordinary flowers came into existence : namely, the 

 calyx, composed of sepals ; the corolla, of petals ; and the stamens, 

 consisting of the filament and anther. 



There seems little doubt that the calyx represents the last leaves 

 of the flowering shoot. These decrease in size and form bracts, and 

 finally two, three, four, or five constitute the calyx, arranged in a 

 whorl through the suppression of the fiower-stalk to form the " floral 

 receptacle." 



Transitions between leaves and bracts are common enough. Thus 

 the Hellebore shows how the petiole of the leaf becomes an oval, pointed 

 bract, while in the Buttercup the bracts represent the dissected blades, 

 the petioles being suppressed. 



That a calyx with separate sepals originated from bracts appears 

 to be seen in the fossil Cycadeoidea as well as in Casuarina, and especially 

 the Alder [Alnus] ; for the male flower of this consists of four " bracteoles 

 or sepals," as Hooker calls them, including four stamens; whereas 

 in the Birch [Betula), the only ally of the Alder, there is a single bract 

 with two stamens, forming each flower. 



With regard to the origin of the corolla, this appears to be due 

 to the response to insect agency ; for the first indication of the appear- 

 ance of petals is in the modification of the anthers into nectaries, 

 and thence into petals, the filament being suppressed, as occurs in 

 the Buttercups, e.g. Ranunculus auricomus ; the anther is open above, 

 the partition between the two cells is arrested, and the resulting tubular 

 or cup-like structure becomes a " nectary " or " honey-pot." 



To form a petal, one side of the tube remains short, while the 

 other — i.e. the outer — enlarges into the lobe we call a petal. 



In the Hellebores and Winter Aconite the nectaries remain as such. 



