512 BOTANICAL GAZETTE [JUNE 
which is therefore triarch. Farther down, the strand which has 
come from the first leaf forks, and the root becomes tetrarch. 
Older seedlings were not examined, but probably further forking of 
strands occurs, making the root eventually polyarch. 
In thus describing the course of the vascular strands, the 
sequence given seems to be that of their formation, although we 
did not observe the procambial strands. If this be true, they are 
laid down first in the growing cotyledon and first leaf, become 
united later in the cotyledonary plate, which then gives rise to the 
root poles. This means that the vascular strands do not determine 
the development of the structures of the seedling, but that the 
primordia of cotyledon and first leaf start and the vascular strands 
are laid down in the growing organs. In other words, growing 
primordia determine the vascular strands rather than the reverse. 
If the cotyledonary sheath is very massive, vascular bundles may 
be laid down, and these may have no connection with other bundles. 
In the cotyledonary sheath of Doryanthes Palmeri, which is very 
thick, a vascular strand was found at a point opposite the massive 
cotyledon. This solitary strand had no connection with the coty- 
ledonary plate, or with any other strand, beginning and ending 
blindly in the sheath. The inference seems evident that vascular 
strands are secondary structures, whose appearance is dependent 
upon the character of the primary structure, and or of no 
phylogenetic significance in the seedling. 
It is noteworthy also that in the seedling of Agapanthus there 
is no stem primordium. In fact, the stem is a very belated organ, 
not having appeared in the seedlings under investigation. When 
and how a stem primordium is organized later was not seen. Cer- 
tainly in the well developed seedlings of Agapanthus there is no 
stem primordium that gives rise to lateral leaf primordia. All of 
the meristematic tissue is involved in cotyledon and leaf-formation. 
The structure of the dicotyledonous seedling of Agapanthus 
shown in fig. 1 is indicated by the series of transverse sections given 
in figs. 15-28, and by their diagrammatic reconstruction in fig. 29. 
The seedling is of the same age as the monocotyledonous one. 
The two cotyledons are the same length, but one is slightly thicker 
than the other (figs. 15 and 16). The cotyledonary sheath extends 
