i 16 Ethel Sargant , 
placed near the two foci (fig. 2). Lower down they are displaced a 
little from this symmetrical position by the formation and growth 
of the plumule (fig. 3). This displacement, too slight to be 
considered significant, makes itself felt into the transitional region 
(figs. 4, 5). The transition takes place, however, in a perfectly 
symmetrical way. Each xylem group gives off three branches of 
protoxylem, and each group of phloem-divides into two (fig. 5). The 
four lateral branches of protoxylem fuse in pairs, and the result is a 
tetrarch root-stele (fig. 6). 
Later observations have strengthened my conviction that the 
vascular characters observed in the seedling Anemarrhena are 
primitive. The same structure—with one trifling exception, to be 
considered later—is found in Alhuca , among the Scilleae, and a very 
similar vascular system occurs in seedlings of the neighbouring 
genus Galtonia. All the other forms which I have examined from 
the Scilleae can be derived from these. For example, the very 
distinct structure characteristic of Hyacinthns and some other 
genera is connected with the Albuca type by a number of inter¬ 
mediate forms found within the single genus Muscari. Further, 
within the genus Scilla we have the Hyacinthns type leading on to a 
form closely resembling that of Lilium and other Tulipeae. Thus it 
appears that the Anemarrhena type is the starting point of at least 
four central tribes within the Liliaceas, and it must therefore be a 
form of some antiquity among Monocotyledons. 
So early as August, 1900, I find a paragraph in my notes 
suggesting that the two bundles of the cotyledon and hypocotyl of 
Anemarrhena may be traces representing the two distinct cotyledons 
of some ancestor. If this view is justified, the “cotyledon” of 
Anemarrhena, and probably that of all Monocotyledons, must be 
equivalent to both the seed-leaves of a Dicotyledon. 
This hypothesis gained in importance and probability as the 
primitive character of the Anemarrhena type became apparent. 
Soon after cutting Albuca and Galtonia , I resolved to examine some 
Dicotyledonous seedlings for comparison. The point of attack was 
decided by the well-known monocotyledonous affinities of the 
Nymphaeaceae and Ranunculacete, and while collecting material 
from both families I was so fortunate as to meet with the detailed 
work on Ranunculaceous seedlings published by M. Sterckx 1 . The 
resemblance in vascular symmetry between such forms as Eranthis 
on the one hand and Anemarrhena and Albuca on the other, was 
1 R. Sterckx, Mem. de la Soc. Royale des .Sciences de Ijege. 
3ierae ser., tow. ii., 1899. 
