84 



The Journal of Heredity 



the compound leaf of the wahiut family 

 appears to have been developed, not by 

 subdivision of the original simple leaf 

 but by reduplication of the foot ele- 

 ment, with its stipular elements, to 

 form the several joints of the rachis 

 and the corresponding pairs of pinnae. 

 The number of pinnae is quite variable 

 in this family, among the leaves of the 

 same trees, with extreme reductions 

 to one or two pairs of pinnae in the 

 bud-scales and small adjacent leaves. 

 Reduplication of stipules occurs in 

 other families, including the ^Malvaceae. 

 Some varieties of okra have four or 

 more stipules, and the involucres in 

 cotton and related genera may have 

 been developed through reduplication 

 of stipules. Although the usual forms 

 of compound leaves do not appear in 

 the Malvaceae, a complex structure 

 may be indicated by the radial lobing. 

 In the walnuts and hickories, at 

 least, we are brought to consider that 

 the major part of the leaf, up to the 

 insertion of the terminal pinnae may be 

 derived from the primitive foot ele- 

 ment, with only the terminal pinna and 

 its petiole representing the blade ele- 

 ment of a simple-leaved plant. Con- 

 sidering that the blade of the primitive 

 leaf may have arisen as a specialized 

 median appendage or lobe of the foot, 

 the stipules may be reckoned as lateral 

 lobes of the same foot element. Thus 

 the foot would appear as a single pri- 

 mary element with one median and two 

 lateral lobes, developed as blade and 

 stipules. In some families only the 

 blade has developed foliar, assimilative 

 functions, in others the stipules also. 

 In the walnut family the stipules as 

 well as the median blade element have 

 had a foliar development, no doubt at 

 first merely supplementing the blade 

 element, but eventually over-balancing 

 and replacing the blade. 



Nature of the Intermediate Leaf- 

 Forms 



A general tendency in the later evo- 

 lution of the compound leaves is to 

 equalize the terminal and lateral pinnae. 



The terminal pinna, instead of being 

 much larger and broader than the lat- 

 eral pinnae, as in the primitive and in- 

 termediate leaf-forms, is reduced to 

 nearly the same size as the lateral 

 pinnae. Indeed, the reduction of the 

 terminal pinna often is carried to the 

 stage of complete suppression, as oc- 

 curs in some of the pecan varieties, 

 and in some species of walnuts, includ- 

 ing Jiiglans caUfornica, the parent of 

 the remarkable "oak" walnut mutation. 

 In this the terminal pinna is regularly 

 developed and the lateral pinnae re- 

 duced to two pairs, or replaced by 

 small stipular organs like those of the 

 bud-scales. (See Figures 13 and 14.) 



When the terminal pinnae are re- 

 duced to the same size as the others, 

 their only distinctive features are bi- 

 lateral symmetry and the possession of 

 a petiole, but occasionally the lateral 

 pinnae also are stalked, as observed at 

 Sacaton, Arizona, in many leaves of 

 the Frotscher pecan. This may be con- 

 sidered as a further step toward equal- 

 izing the pinnae, with the lateral pinnae 

 beginning to show a character that pre- 

 viously was confined to the terminal 

 pinna. Likewise, the development of 

 the rachis by intercalation of more 

 numerous joints and additional pairs 

 of pinnae may be considered as a re- 

 flection or imitation in the leaf of the 

 jointed structure of the stem. 



The evolutionary interest of such 

 equalizing tendencies among the differ- 

 ent parts of plants was first recognized 

 by Leavitt, who considered such char- 

 acters as having been transferred from 

 one part of a plant to another, by a 

 l)rocess which he called homoeosis. 

 But if it be supposed that the charac- 

 ters exist potentially in all the parts of 

 a plant, the idea of a transfer seems 

 unnecessary, since the apparent trans- 

 fers may result from changes in the 

 expression of the characters. Such 

 changes would account not only for 

 different degrees of expression but also 

 dift'erent combinations with other char- 

 acters. In the words of East : "One 

 must realize that each gene has many 



