Oct., 192 1 ] BECHTEL — FLORAL ANATOMY OF URTICALES 
399 
i^). The anterior bundle (figs. 14, 16, 17, 0 + i) continues for some distance 
and then separates into two unequal strands, one passing into the anterior 
side of the pistil and the other, the larger, passing into the ovule (figs. 14, 
18-20, 0,1), The former strand soon vanishes in the lower third of the ovary 
wall (fig. 14, i). This is evidence of an abortive anterior carpel. 
The ovule and its vascular supply again offer opportunity for specula- 
tion. After the anterior carpel strand separates from the single anterior 
strand, the main supply passes horizontally in an ascending-posterior direc- 
tion through a long stocky funiculus into the ovule (figs. 14, 18-20, /, 0). 
The position of the ovule suggests the reduction of an axillary placenta 
which bore ovules in a pendulous position. The ovule is past the midway 
stage between that of a pendulous ovule as in Ulmaceae and Moraceae and 
that of a basal ovule as in Boehmeria cylindrica. In fact, very little reduction 
in funicular tissue in Laportea canadensis is necessary to duplicate in position 
that of the ovule of Boehmeria (compare PI. XXII, figs, i and 14). The 
same line of reasoning is suggested upon comparing the ovule supply of 
Boehmeria cylindrica and Urtica gracilis (PI, XXI, fig. i; PI. XXII, fig. i), 
namely: the bundle in its indirect route to the basal ovule of Boehmeria 
would require little reduction to duplicate the direct supply to the basal 
ovule of Urtica. 
Discussion 
Ulmaceae. The gross floral morphology of the six species and the 
detailed anatomy underlying it have been presented above. The latter 
reveals features which warrant the disuse of the descriptive term "simple" 
for the flowers: five sets of organs or vestiges of organs; variableness in 
number of parts in a whorl; zygomorphy, which is constant; and the fusion 
of like and unlike parts. These present a decidedly complex condition. 
The flowers of the genus Ulmus have a perianth "cup" upon the edge of 
which the perianth parts and stamens have been considered perigynously 
inserted. Baillon (i) considered this to be the condition. Anatomical 
work reveals that this is not the case for the following reasons. First, the 
vascular supply to the stamens and to the perianth parts arises separately 
from the stele of the pedicel ; the former passes off from the floral axis con- 
siderably above that to the latter in U. fulva, U. campestris, and U. scabra 
(PI. XVII, figs. I, 5, 8, b)y but approximately closely in U, americana and 
U. racemosa (PI. XV, fig. 3; PI. XVII, fig. 3, h). Second, the tissues em- 
bodying the perianth and stamen traces through the perianth "cup" are 
separable by the distinct difference in cellular structure and by a line of 
demarcation. These differences in the parenchymatous tissue are col- 
lateral and continue into the perianth lobe and stamen respectively. The 
line of demarcation indicates an adnation of the tissues of the perianth 
lobe and stamen. Here is good evidence that the "cup" is the fused bases 
of floral envelopes and stamens. Third, the lobes of the perianth are 
variable in length; their size is not constant, which is a character not un- 
