704 



THE AMERICAN NATURA LIST [Vol. XLI 



in Fig. 7b. If the process of compounding proceeds further, a 

 smaller secondary leaflet will be cut off opposite the one on the 

 lower border of the primary leaflet (Fig. 7c). Thus the basal 

 primary leaflet becomes pinnate and develops further in the basi- 

 fugal manner. This order of leaflet formation is seen not only 

 in the elder, but in Aralia nudicaulis (Figs. 8a-8c) and in Cicuta 

 maculata (Figs. 9a-9b). It is of widespread occurrence. 



An unusual exception to the basal formation of secondary leaflets 

 is seen in many leaves of Bidens frondosa (Fig. 10). In this 

 species the secondary leaflets are usually on the upper margin of 

 the basal leaflets. They may become matched by leaflets on the 

 lower margin, and sometimes the leaflet on the lower side is formed 

 first. Frequently in Sambucus canadensis the secondary leaflets 

 first appear on the distal sides of the basal leaflets, as in Bidens, 

 but usually they develop on the basal side, both in Sambucus 

 and in most of the species examined. 



The development of the pinnate leaves of the rose, as described 

 in the previous paper, is so different from that of other pinnate 

 leaves as to require further study. Eichler ^ has classed with the 

 rose, as basipetal in development, the leaves of Sanguisorha 

 officinalis, Poterium sanguisorha, Potentilla anserina, and " probably 

 all potentillas with compound and divided leaves." Tlie basi- 

 petal nature of the palmate leaves of Potentilla canadensis is ob- 

 vious. In "Gray's Manual" they are described as "ternate but 

 apparently quinate by the parting of the lateral leaflets." Fre- 

 quently they develop seven leaflets without lobation of the central 

 leaflet. In Potentilla fruticosa, however, the central leaflet shows 

 various degrees of indentation, and if one may judge from mature 

 leaves, basifugal growth occurs. The entire leaf is pinnate. 

 Potentilla anserina also shows lobed terminal leaflets.^ It is 

 possible in these forms that the proximal leaves are added basi- 

 petally but they are not coiniected with tlie stii)ules, and lobed 

 proximal leaflets were not observed in the y)lants examined. 



11) was shown in two leaves among four lumdred and lifty. 

 1861. 6() pp. 



' Goebel agrees with Eichler in considering Potentilla anserina as basiix;tal. 



