212 BOTANICAL GAZETTE [SEPTEMBER 
leaf as pseudo-girdling, whose cause is to be sought in the accident 
of its development. When the inner face of a leaf. encloses a mass 
whose diameter is less than 3™™ (fourth leaf in fig. 12), the vascular 
bundles of that leaf are all vertical; but when the enclosed mass, by 
its enormous radial growth, has reached a diameter of 5-8™™ and 
comprises two or more developing leaves (second leaf in fig. 12) the 
base of the enclosing leaf enlarges its inner face accordingly. The first 
stage in the enlargement consists in an increase in the number of 
cells, but the second in a horizontal elongation of these (text jig. 1). 
While this is going on the older part of the central vascular cylinder 
is also increasing its diameter, separating farther and farther the 
original positions of the bundles of the leaf. As a result, the marginal 
traces gradually elongate as they are drawn more and more from the 
vertical position, and their upper parts stretch outward in the direction 
which the leaf takes. These facts have been observed repeatedly; 
whether they are an adequate solution of the problem of the cause of 
girdling, as Matte thinks, I am unable to say. 
In seedlings with three or four leaves, the stem bundles (#, jig. 19) 
branch repeatedly, and many of the branches reunite to form a sm 
number of traces. Each of the remaining ones now forks once, the 
larger member in each instance going to the leaf as before, the smaller 
one continuing in the axis. Thus, even at this early stage, there 1s 
present the sympodial stem described by Miss Sarr (12) for older 
plants. 
The number of strands entering successive leaves Was seen {0 
increase, sometimes with great regularity. In the plant “ee 
dissection is represented in figs. zz and 12, the number of traces 
was increased by one in each successive leaf, from the cotyledon 
with three, to the fourth leaf with seven; but the increment W% not 
so constant in all the plants observed. 
Within the leaf base and-in the petiole, the bundles branch and | 
anastomose freely. There is no real Q in these first leaves; the 
bundles are arranged simply in an open arch. Text fig. 2 is 
of a foliage leaf which was the second lateral organ of the io 
that is, it was preceded by only one scale leaf. The scale leaf a 
four bundles; this leaf has five (a, 8); but just above the stipules 
number is reduced to four by the approximation of two of them (0) 
