26 Rhodora [FEBRUARY 
Characters may pass from members of one kind to those of a 
morphologically different kind. Morphic translocation seems usually 
to be conditioned not by homology, but by similarity in the develop- 
mental substratum. The rudiment of an ovule, which is an emer- 
gence of the placental tissue, is much like that of the Droseraceous 
tentacle; this I can attest from a study of the development of the 
latter organ. It is also like the beginning of the leaf at the summit 
of the stem. The partly developed nucellus is a mound of embryonic 
tissue not unlike the growing stem tip. When vegetative impulses 
invade the flower the formative forces seize upon the ovular or 
nucellar rudiments and mould them into vegetative organs, such as 
are normally produced in the vegetative parts from similar originals. 
‘The ovules, or rather their substitutes, then appear as tentacles, 
leaves, etc., even as whole leaf-bearing shoots. 
I have not here discussed the merits of the questions at issue, but 
have criticized the arguments. However, in the case of the enlarged 
ventral canal cell, it seems to me that reversion, in any true sense, is 
extremely unlikely. To find an archegonium-like organ in which 
more than one gamete was a normal condition, we should have to go 
back over an enormous stretch of history, certainly to the predeces- 
sors of the lowest Hepatic, perhaps into the Algz. I know of no 
clear indication that characters remain latent for such lengths of evo- 
lutionary advance. Characters clearly Pteridophytic do not appear 
anomalously in Angiosperms, nor Bryophytic anomalies, so far as 
I am aware, in Ferns. The approximation of the ventral canal cell 
to the egg in size and structure is probably brought about in the 
way already suggested: a character which in the course of normal 
events has been evolved in one member, the egg, has thereby become 
potential for another member similarly conditioned and of like form 
in the inceptive stage, and may abnormally become actualized in 
that other member, the ventral canal cell. Zhe precedents for mon- 
strosities may thus often be sought in contemporary normal parts, rather 
than in ancestral conditions. 
MORPHIC TRANSLOCATION IN EvoLUTION. — While organic chaos 
would result from frequent substitutions occurring in the normal con- 
structive sequence of ontogenetic development and while great trans- 
locational capacity would be antagonistic to adequate permanence of 
type, yet within certain limits the principle might operate without 
disastrous results and might even, by introducing new combinations 
