22 



DEVELOPMENT OF SIUM CICUTAEFOLIUM. 



also the figures upon which Jackson and Cushman base their conclu- 

 sions, I am convinced that with possible rare exceptions the new char- 

 acter added to a leaf having a single blade and an entire margin consists 

 in an indentation or an incision rather than an outgrowth, and that, in 

 general, increased complexity is brought about by an increasing number 

 and depth of these incisions, which do not and can not occur at the 

 base, but must occur above it. That part of the base below the lowest 

 incision is characterized by an entire margin and is " primitive," there- 

 fore, in the same sense as is that part of the apex above the highest 

 incision. Just as Cushman (1903, p. 244) noted that emarginate apexes 

 may develop from acute ones in Astragalus adsurgens, so in cases like 

 that of his Arabis albida, in which truncate and cordate bases occur in a 

 series beginning with attenuate bases, these forms are late modifications 

 of a tapering base, due to the excess of marginal growth as compared 

 with that of the mid-rib. Although the basal margin is thus distorted 

 from its original direction, it retains its unbroken "primitive" character. 



The figures which 

 illustrate Cushman's 

 papers show these 

 facts so clearly that 

 it seems strange that 

 he should not have 

 recognized them. 

 Particularly in Sib- 

 bald'wpsis{PotentiUa) 

 tridenata, Artemisia 

 stelleriana, and San- 

 gnisorba (Poteriuui) 

 canadensis is it diffi- 

 cult to harmonize his 

 figures with his state- 

 ment that new charac- 

 ters are produced at 

 base. In the first two 

 of these plants the new 

 character, as I under- 

 stand it, consists in the appearance of two small indentations very near 

 the apex. In Sanguisorba canadensis the leaf is pinnate, and in the 

 later stages of the senescent series each proximal leaflet presents in each 

 leaf a peculiar inequilateral form, though the number of leaflets de- 



FiG. 10.— Diagram to test by the law of 

 periodicity tlae serial homology of 

 leaflets in a pinnate leaf. Horizontal 

 lines represent the rachises of the 

 successive leaves of a stem of Siuin 

 cicutae folium . Nodes assumed to be 

 homologous are connected by curves. 

 The proximal leaflets assumed to be 

 homologous. 



