Syncotyly and Schizocotyly. 813 
men’s seed. A considerable number of seedlings of each species was 
examined, the specimens ranging from dicotyls to tetracotyls, through all 
intermediate stages. The structures are very uniform for all the species, 
and closely resemble those described for Phacelia , so that I need not enter 
into details. The following is a summary of results. 
In all the species the root is diarch in the dicotyl, and, since the seed- 
lings are very small and slender, the transition is high, root-structure being 
found to within one or two millimetres of the cotyledons. Each cotyledon 
takes the strand derived from a single root-pole and a pair of half-phloems. 
Splitting of a cotyledon or an increase in number of cotyledons involves 
the increase of the number of xylem-poles, this being almost always accom- 
plished by the addition of one or more de novo. 
Triarchy is frequent in tricotyls and seedlings of higher grades of 
schizocotyly. Tetrarchy is occasionally found in tetracotyls. 
Lepidium SATIVUM (Cruciferae). 
The cotyledons of Cress are pinnately three-lobed (Fig. 36). It was 
thus of interest to ascertain if schizocotyly occurs in this species, and, if so, 
what effect it has. Large sowings were made, and among them were found 
three tricotylous seedlings. Each of these seedlings had three equal and 
similar trilobed cotyledons. One of these was pickled and microtomed ; the 
other two were grown to maturity in pots, and their seeds (self-pollinated or 
crossed by one another) were sown. The offspring were almost entirely 
dicotylous, but after some search four complete tricotyls were found — • 
about 0-5 per cent, of the whole number of seedlings. 1 These four seedlings 
were also studied anatomically. No other abnormalities of the class we are 
at present concerned with appeared in the seed-pans. 
Dicotyls. The root is diarch, and the transition is high. At the 
cotyledonary node the ‘ double bundle ; is practically tangential. On 
entering the cotyledon petiole it becomes a single collateral strand, giving 
off a pair of lateral bundles which pass into the lateral pinnae (Fig. 36). 
Tricotyls. Three of the five tricotyls (A, B, C) were triarch throughout, 
the behaviour of the vascular strands being strictly analogous to that found 
in dicotyls. 
One tricotyl (D) showed a diarch xylem in the root; just above the 
collet a third protoxylem appeared laterally but towards one end, thus not 
dividing the phloem ; the new xylem gradually increased in size until it 
equalled the two original ones, and a new phloem appeared. The upper 
part of the hypocotyl was constructed on the triarch plan. 
The fifth seedling (E) showed near the root-tip a xylem-plate with one 
protoxylem at one end, and two somewhat unequal ones at the opposite 
1 It therefore seemed clear that a rich ‘ intermediate race ’ was not obtained, and the experiment 
was discontinued. 
