1905 ] Leavitt, — Translocation of Characters in Plants 29 
sharp and rigid; the stipels, though weaker are rigid and sharply 
pointed. In most other Papilionaceae where both kinds of organs 
occur the stipules and stipels correspond closely in being slight, fili- 
form structures. In Sambucus canadensis the stipels are usually 
terete — occasionally flat and foliaceous — and capitate, with a small 
glandular depression at the end; the stipules which I found on vig- 
orous suckers were of the same description. 
While we may not be able to determine with certainty whether 
the present peculiarities of the stipules in these species were taken 
on before or after the appearance of stipels, yet the way in which the 
two ranks of members accord in detail as they run the gamut of 
forms is indicative of some fundamental relation between them. I 
shall define the nature of a possible or probable relation presently. 
As to absciss-layers, it is first to be said that the casting of the 
leaves of deciduous perennials is an active process, and that the 
result is achieved by the agency of definite cellular structures. 
Shortly before the time of defoliation active cell-division begins in 
the tissue at the base of the petiole, where separation is to take 
place, and a plate of cells is organized. The middle single layer of 
the plate finally disintegrates and the leaf drops off (when the fibrous 
strands break) leaving a clean scar. 
If one will examine any compound leaf at the time of leaf-fall, he 
will probably find that as the petiole is disarticulated from the stem, 
so the leaflets are disarticulated from the rhachis. Eventually the 
leaf not only falls off, but falls to pieces. Indeed the leaflets often 
depart while the leaf stalk and rhachis still stand upon the stem held 
by the fibrous elements, which are not cut in abscission. As far as 
my observation goes — and I have examined numerous species and 
groups, ranging widely both taxonomically and geographically — the 
casting of leaflets is universal in deciduous, compound-leaved plants. 
The appearance of the leaflet scars, when compared with leaf scars, 
at once suggests that the defoliating process is the same in both 
cases; and the exact researches of von Mohl! leave no doubt on 
this head. The abjection of leaflets is the same structually positive 
process as that of leaves. At the base of the petiolule cell divi- 
sion sets in, resulting in the organization of an absciss-layer. 
We see, then, that in the evolution of compound leaves two kinds 
! Bot. Zeit. 18:274 (1860). 
