706 



THE AMERICAN NATURALIST [Vol. XLI 



Twenty-seven hinidred leaves of Rosa liicida yielded none with 

 a lobed terminal leaflet, but basii'iiiral n;r()wth was sufj^-ested by 



ary leaflets attached to the distal j)air of primary leaflets were 

 observed (Fi,s^. In Sambueus, .\ralia, and (^ieiita, the oldest 



leaflets are the ones which ^/i\v rise to secondary leaflets, and they 

 are conscfinently found toward the base of tlie leaf. Tiu^r distal 

 position in the rose may be correlated with basipctal development. 

 In the celandine, however, in which orowth seems ch>arly basifugal, 

 the distal leaflets jiroducc secondare leaflets as in the rose (Figs. 

 13aandb3b). 



Among the twenty-seven hundred rose leaves there were none 

 with the |)roximal leaflets lobed. One pedate leaf was found 

 (Fig. 1 2d ) t()g(>thcr ^\ ith scn eral forms like that in Fig. 12e. These 

 suggest that leaflets cut oil" from the basal pair may be carried 

 down the j)etiole as should occur in a truly basipetal pinnate leaf. 

 However, lobation of the yn-oximal leafl(«ts of a tcrnate leaf leading 

 to the production of a pinnate leaf has never l)e(>n found by the 



