METAMORPHOSIS. 137 



to the original staminal nature of tlie septils, to fuse 

 with the "cohimn." Hence this may be included under 

 " partial staminody." 



The cases of staminody, somewhat uncommon though 

 they are, certainly support the view that the sepals 

 have been derived from stamens, and they therefore 

 represent reversions. 



Cabpi';i,lody. — This is perhaps somewhat more fre- 

 quent. Laxton observed a flower of the garden pea 

 {Pisam sativum) in which some of the segments of the 

 calyx of a second flower, formed by proliferation within 

 the normal one, bore imperfect ovules on their margins 

 and their tips were prolonged into styles. It is not 

 rare for the sepals of the tulip and crocus to become 

 partially carpelloid, bearing ovules on their margins ; 

 in these cases all transitions have been found between 

 carpels and sepals. These phenomena cannot be held 

 to be cases of direct reversion. But on the view that 

 the calyx has been derived from stamens, and that, 

 moreover, as we shall see later on, the latter have a 

 very close affinity, and are often interchangeable, with 

 carpels, they maybe regarded as case? of very indirect 

 reversion. 



(2) Corolla. 



Phtllody and Vieesoenoe. — In Primula sinensis the 

 petals showed but a slight change in this direction, 

 being slightly enlarged and their upper surfaces green. 

 In the primrose (P. valrfarls) flowers were seen whose 

 petals were of the normal size but quite virescent : all 

 of a yellowish-green colour, the stamens being, in 

 correlation therewith, degenerated (PI. XL, fig. 4). 

 Flowers of the honeysuckle {Lonicera Periclymenwn) 

 and florets of the Heleninm above described, have been 

 affected in the same way. In Begonia Pea reel one 

 petal was half-, the other completely virescent. In 

 Gattleya Loddigesii all the perianth -leaves, except the 

 lip, were in the form of narrow, virescent leaves, the 

 flowers so affected being greatly dwarfed. Much the 



