BOTANICAL SUBJECTS. 237 



What concerns me in these pages is a broad comparison of 

 parts of plants from whatever di\4sion of the vegetable kingdom 

 they may come, and I do so, because I accept the creed that all 

 vegetable structures, however complex, have evolved from simple 

 unicellular plants. 



In this transformation of leaf to stamen, all manner of varia- 

 tions occur. In the curling of leaves when young we probably 

 get a hint of the origin of extrorse and introrse anthers. In the 

 Polygonacece, the young leaves are curled extrorsely ; while in the 

 Viola, they are curled introrsely. 



In the variations which stamens, undergo, we find that, not 

 only are the stamens separated from the pistils into distinct 

 unisexual flowers, but also into fUstinct trees. The anther is 

 attached to the filament in various ways, and it opens to let out 

 the pollen in one of many ways. In some plants it opens by a 

 sHt, in others by pores or valves. What I said regarding the 

 dehiscence of the ripe ovary would apply to the dehiscence of the 

 anther. In some stamens the anthers are turned inwards, in 

 others outwards. Some are on the top of the filament, some on 

 the sides, and so forth. 



For all these little differences botanists have invented distinct 

 names, which may have some value in shortening descriptions, but 

 which are certainly bewildering to beginners. 



The one-celled anther, such as that of Gomphrena, Mallow, 

 and others, is not difficult to conceive as having been evolved out 

 of a leaf-blade with incurved margins, such as that of Viola and 

 others. This incurving of the margins is repeated in the carpels 

 of many plants. 



Then, if the surface of this leaf be covered ^vith globular hairs, 

 such as are plainly visible on the young leaves of Chenopodium 

 petiolare, we shall have the reproductive hairs, which we call 

 pollen in the stamen. 



The two-celled anther seems at first more difiicult to under- 

 stand, as to its origin from a leaf -blade. If, however, we examine 

 the two-celled anther of the Hornbeam, we soon become aware 

 that this stamen may be made up of two one-celled stamens joined 

 together bi/ fusion of their //laments. Further fusion would 

 reduce them to the two-celled anthers we see everywhere. 



Fusion, as I have shown elsewhere, has been one of the great 

 modifiers of organs, and we should not forget the possibility of 



