814 
Compton . — An Anatomical Study of 
end ; two phloem bands were present. The xylem gradually became sym- 
metrically triarch, a new phloem appearing in between the two adjoining 
protoxylems. Doubtless, if the seedling had been a little older, diarchy 
would have been found at the root-tip ; but as it was there was no means of 
knowing the manner in which the three protoxylems could have joined on 
to the diarch plate — whether both or only one of the neighbouring ones 
would be prolonged downwards. 
There is nothing in the external morphology of these seedlings to 
suggest schizocotyly : the vascular system, too, is either trimerous, or the 
third xylem arises in the ‘ subsidiary ’ manner (in D and perhaps also in E). 
It may be that this is an instance of true meristic variation — ‘absolute 
Vermehrung ’, as Winkler 1 calls it. Winkler, indeed, remarks that this 
mode of increase in the number of cotyledons is most obvious in seedlings 
with divided cotyledons ; but the argument appears to be fallacious, and at 
present, owing to the small numbers observed, the question must remain 
unanswered. 
Lotus corniculatus (Leguminosae). 
The anatomy of dicotyls of these species has recently been described. 2 
The root is normally triarch, each cotyledon receiving an equal share of the 
root xylem, viz. one whole protoxylem and half of the intercotyledonary 
protoxylem — the latter, however, being somewhat reduced in bulk. It is 
therefore a curious problem as to what will be the structure of tricotylous 
and hemitricotylous seedlings. One of each class was found and studied, 
with the following results : 
Tricotyl. The root was triarch, as in dicotyls, and the whole transition 
was carried out on this plan of symmetry. Each cotyledon took a single 
triad derived from one root-pole. Thus the behaviour of one of the three 
root xylems was changed ; instead of dividing itself between two cotyledons, 
as in a dicotyl, it served a single cotyledon. 
Hemitricotyl. One cotyledon was split for about one-third of its 
length. The root was triarch, and the whole transition was carried out 
exactly as in the dicotyl. The base of the split cotyledon contained the 
usual ‘ double bundle * with two phloems (Fig. 37). A little distance along 
the petiole one of these phloems gave off a branch towards the median 
plane, so that a xylem bundle with three dorsal phloems was produced 
(Fig. 38). Further along the lamina this vascular complex divided into 
two equal halves, each consisting of a xylem with two phloem groups — in 
fact a ‘ double bundle ’ (Fig. 39). These two double bundles supplied the 
two lobes of the bifid cotyledon. 
Thus the hemitricotyl yielded no hint as to the mode of origin of the 
tricotylous type of anatomy. 
1 Winkler (’84), p. 39. 
2 Compton (’12), p. 38. 
