1905] Leavitt,— Translocation of Characters in Plants 27 
of old characters, contribute to evolutionary progress. But before 
the translocational property can be admitted as an evolutionary factor 
it must be shown to be transmissible from generation to generation. 
Two of the cases described early in this paper are in point. Mr. 
Fernald states that he found the variety of Drosera rotundifolia 
which he has called comosa, in great abundance, covering the bog 
where it grew. When a trait is found thus in very numerous indi- 
viduals it must be entered in the evolutionary account book. The 
pentadecandrous Saxifrage seems also to be a fixed form. Sterns 
judged it to be propagated by seed; and Mr. Sears has written me 
that he has found it — in not a little quantity, as appears from the 
amount of material sent to me —at a certain station repeatedly since 
1886. The experience of gardeners in the propagation of “double” 
flowered races of plants, in which stamens have turned to petals, is 
well known. Such forms of Begonia, the Stock, Hollyhock, Columbine, 
etc., may be raised from the seed. I may also adduce the experi- 
ments of De Vries? with an aberrant Zimaria vulgaris. The flowers 
were peloric. The spur character had passed from the anterior petal 
of the gamopetalous corolla to all the other petals and the corolla 
tube had thus become s-spurred. In three generations from the 
seed of peloric plants De Vries found, out of 119 individuals, 106 
fully peloric. The form was, therefore, in a high degree, although 
not fully, true to seed. Changes of organization of the kind here 
being considered are therefore inheritable, and call for a place in the 
scheme of evolution. 
TRANSLOCATION OF CHARACTERS IN COMPOUND LEAVES, FOLLOW- 
ING MODIFICATIONS ARISING IN THE COURSE Of NomMar EvoLu- 
TION. — With relation to the main theme of this paper I wish now 
to point out certain aspects of compound leaves. Compounding of 
the leaf is found in many divisions of Dicotyledons from the most 
primitive, as Juglandaceae, to the most advanced, as Caprifoliaceae. 
The compound-leaved groups are separated by entire-leaved allies in 
such a way that we know compounding to have arisen repeatedly and 
independently. In families which include some species with entire 
leaves, others with leaves lobed, parted, etc., and finally some with 
leaves fully compound, we see, as it were, the compounding in proc- 
ess of development. It is quite certain that this process has been 
! RHODORA 7:9. ? Die Mutationstheorie 1: 562. 
