THE MAKING AND UNMAKING OF FLOWERS. 



117 



a difficulty, because no strictly intermediate step is known. There is 

 a genus, a single type of a family called the Beefwocds of Australia 

 (Casuarina), which is a solitary genus, a relic of a bygone ancestry, 

 which has one or two points of resemblance, but one can hardly regard 

 them as indicating affinity. Thus, just as Gnetum has a single stamen 

 within a coherent perianth, so, too, has Casuarina. As the fruit of 

 Welivitschia is invested by a close-fitting perianth, so is it in Casuarina. 

 But while the latter has a perfect pistil, no member of the Gymnosperins 

 has any at all. 



Our Sweet- Gale [Myrica) has certain affinities with, or at least 

 resemblances to, Casuarina. 



There are some other members of the "Incomplete" division of 

 Dicotyledons, which we may feel inclined to regard as primitive types, 

 on account of the extreme simplicity of their flowers, and also because 

 they are isolated genera. Thus Willows (Salix) and Poplars {Foimlus) 

 constitute an entire Order. In the former (fig. 39) there are two or more 

 stamens in the axil of a bract for a male flower ; while the female has 

 only a pistil composed of two carpels. Moreover, Poplars are among the 

 earliest of flowering plants known to geologists. 



All these and many others afibrd no true connections with Gymno- 



sperms, as they are provided with perfect pistils, having styles and 

 stigmas. Hence the links are still wanting to connect these groups. 



We must now suppose that Nature has succeeded in making a calyx, 

 at least, within which are either stamens or a pistil, or both together. As 

 every part of a flower is homologous with a leaf, and as leaves are not 

 joined together in any way, we must look for a flower having all its parts 

 separate. Thus a Buttercup, and other members of the same family, will 

 answer this condition. Conjointly with this, I will take the Winter 

 Aconite and two kinds of Hellebore, for in the last two Nature has not 

 yet made a corolla. In the middle of these flowers are several free carpels ; 

 then, numerous stamens ; around them are little honey-secreting vessels, 

 where we should expect a corolla ; and lastly, is a green calyx in Hellebore, 

 and a yellow one in the Winter Aconite. 



Now, what is the origin of these little honey-pots ? A close examination 

 will reveal that they are made out of anthers. They are open at the top, 

 no pollen is formed, but instead, the inner surface secretes honey. In the 

 Winter Aconite especially the outer side is a little longer than the inner 

 (tig. 40, a). If now we compare it with the petals of various species of 

 Buttercup, we soon find transitions in size between those "nectaries," as 



a h c 



Fig. 40. — a. Nectary of Winter Aconite ; 

 6, of Christmas Rose ; c, of Fetid 

 Hellebore. 



Fig. 41. — Transitions from stamens to 

 petals in a double Rose. 



