OF PLANTS. 287 



are not unfolded, but leaves and other lower organs arise 

 instead of them. We thus observe, that the blossoms of the 

 Juncus subverticillatus, when it remains as Junciis Jluitans 

 constantly under water, degenerate into long stem leaves. 

 In the same manner, it is not uncommon, with the Rubus 

 fruticosus, when in dark forests it is deprived of the sun's 

 light, to put forth only leaves instead of blossoms. In the 

 double flowers of Hesperis matronalis, we frequently remark 

 the transition into the leaf form. In the Colchicum autum- 

 nale, not only the blossoms, but even the filaments and pis- 

 tils, have been seen to assume the colour and form of stem- 

 leaves ; (Bernhardi in Homer's Archiv. b. 2.) When the 

 garden tulip is very double, the outermost petals are of- 

 ten mai'ked with green streaks, and even the innermost, 

 which have arisen out of the pistillum, shew the same colour- 

 ing. 



When seeds pass into bulbs, as has been observed in Bui- 

 bine Asiatica and Moi^ea Northianay the same kind of retro- 

 gradation in the process of vegetation takes place, as when we 

 observe that the siliques of Clover degenerate into leaves, or 

 that a pear pushes out leaves (Keith, ii. tab. 9. fig. 12.), and 

 that one flower arises from another ; a mode of growth which 

 in some cases is a law of nature, but which in our Centifo- 

 liae, and in other instances, is a consequence of malformation 

 produced by luxuriant growth. 



415. 



These retrogradations in the process of vegetation, explain 

 the circumstance of blossoms becoming multiplicate, and j>er- 

 fectly double or full. Cultivation stimulates the organs of 

 nourishment, and the instruments of propagation pass into 

 these. Yet there are inferior gradations of this disposition to 

 become double, in which the organs of fructification remain un- 

 confined in their evolution, and in the exercise of then- fimc- 

 tions. When in the Hydrangea of our gardens, the jiarls of 

 the calyx expand, and become of the nature of a corolla, 

 the evolution of the filaments suffers so little by this, that we 

 frequently observe eight of them instead of five. In the t^ame 



