20 Sargant. — T heory of the Origin of Monocotyledons 
Before describing hypocotyls, in which the main cotyle- 
donary traces play a subordinate part, I must mention an 
interesting variant on the form of transition just described 
{Muscari armenaicum A 5 , and others), which is found in 
S cilia sibirica. 
Two main traces and two lateral ones enter the hypo- 
cotyl from the cotyledon. The four xylem strands are con- 
tinued straight downwards into the root, and the protoxylem 
of each becomes external during the passage. Each of the 
four phloem groups divides into two, and the half of each 
bundle unites with the half of the bundle lying next to it 
A 
B 
C 
Diagram III. 
(B, Diagram III). In short, the four cotyledonary traces become 
a tetrarch root- stele according to Van Tieghem’s type 3. 
I have examined no intermediate forms which would suggest 
whether this symmetry is derived from that of an ancestor 
resembling Muscari armenaicum , or more remotely from one 
with the vascular structure which I have called type 5 ( Galtonia , 
Dipcadi). 
In Muscari comoSum , Hyacinthus orientalis , Scilla festalis , 
Salisb., and Ornithogalum sulphureum , more than two lateral 
traces from the cotyledon enter the hypocotyledonary stele, 
and the two main traces form less than half the root-stele. 
Such a possibility was suggested by the structure of seedling 
A 2 in Muscari neglectum. It is worth notice that the main 
traces of Muscari comosum and Hyacmthus orientalis are dis- 
tinguished not merely by their orientation in the sheath, but 
