BOTANICAL SUBJECTS. 63 



splits up into its component parts, which are indistinguishable 

 from the petals proper. This occurs markedly in " Telamonius 

 plenus.^' 



I think we can trace the genesis of the narcissus crown. The 

 " forget-me-not " at the base of each petal and forming a circle 

 round the throat of the flower, has a two-lobed nipple. Sometimes 

 two adjacent nipples fuse into a broader one. Some varieties of 

 Primula polyanthus have a diminutive lobed crown at their 

 throat. The Hippeastrum, as I said, has fringed appendages in 

 its throat ; all these by modification and selection might be coaxed 

 into small crowns similar to that of " Narcissus poeticus^^ and 

 others, which have very small crowns. Compare the crown of the 

 latter with that of the " Glory of Leyden." 



Then the crown of Eucharis, Pancratium^ and others, may be 

 formed by the fusion of the stipule-like appendages of the stamens. 

 They are not only stipule-like, but probably are the homologues 

 of stipules, for we have no right to suppose that, because a leaf 

 has been transformed into a stamen it has lost all power of 

 reproducing stipules. Stamens with stipule-like appendages are 

 found in many plants of various affinities, and they are almost 

 a copy of the stipules of the leaf of Br ay era Abyssinica. In 

 Eucharis and Pancratium it appears that the fusion of the 

 margins of the staminal stipules makes a contitiuous crown. 



Flowers with many petals and many stamens may have origi- 

 nated from \\\Q fusion of two or more flowers, consisting of a much 

 smaller number of segments. The eight-petalled daffodil, the ten- 

 petalled tulip, the nine or ten-petalled " forget-me-not " are examples 

 of what can happen, and broad hints of what may have happened 

 during the long ages and processes of Creation. 



Among the Eosacene, for instance, we have plants A^^th flowers 

 containing only one akene and four stamens, as in Alchemilla, or 

 two akenes and five stamens, as in Agrimonia. If two flowers 

 of these fused, and by doubling the carpels and stamens formed a 

 symmetrical flower, we would have one with two akenes and eight 

 stamens, and another with four akenes and ten stamens. If this 

 doubling happened to go on now and again through the millions 

 of generations that have passed over plants, from Alg« to the 

 higher Phoenogams, it is quite conceivable that we might 

 ultimately obtain a rose, a strawberry, or a raspberry, with an 

 indefinite number of akenes and stamens. 



