Syncotyly and Schizocotyly. 817 
(’02, ’ll). No pure-breeding races of tricotyls were obtained, but ‘ inter- 
mediate races ’ were frequent. 
On the anatomical side schizocotyls have been examined by Leger (’90), 
in Acer platanoides, Gain (’00), in Phaseolus , Guillaumin (’ll), in Ruta 
montana , and Schinus terebintifolins , Lutz (’ll), in Dianthus margaritae . 
Notes on their writings will not be out of place here. 
Dianthus margaritae. A single tricotyl examined by Lutz (’ll) 
showed a triarch structure throughout. No mention is made of the 
anatomy of dicotyls. 
Ruta montana. The hypocotyl of a tricotyl showed a triarch sym- 
metry, but Guillaumin (’ll) does not mention the root structure nor the 
dicotylous anatomy. 1 
Schinus terebintifolius. A few diagrams are given by Guillaumin (’ll), 
but these throw no light on the structure of the root, nor the transition 
in dicotyls nor in schizocotyls. 
Acer platanoides . A full and careful study of a series of schizocotyls 
has been made by Leger (’90) in this species, which regularly produces 
a relatively high percentage of abnormal seedlings. 2 The dicotyl is tetrarch, 
so that Leger’s work is complementary to my own, in which no tetrarch 
seedling is included. 
The mode in which the gradual increase in number of cotyledons 
through schizocotyly is reflected in the structure is somewhat complicated, 
and reference must be made to Leger’s paper for details. The anatomical 
anomaly may be described as a gradual increase in the importance of the 
vascular system in the margins of the half-cotyledons adjoining the line of 
fission. When the fission is deep enough these internal marginal systems 
produce an augmentation in the, number of bundles in the hypocotyl, and 
finally in the root : complete tricotyly produces pentarchy and hexarchy ; 
tetracotyly produces heptarchy (and octarchy is to be expected). 
Phaseolus sp. The variety known as ‘ Haricot beurre nain Mont-d’Or ’ 
produces a high percentage of tricotyls, which were studied anatomically by 
1 The genus Ruta (like other Rutaceae) is of special interest in this connexion. Gerard ( (’71), 
p. 341) states that seedlings of R . graveolens are either tricotylous, or else possess one small and one 
large forked cotyledon : in the latter case the large cotyledon takes two, the small one of the three 
root-poles ; in the former each cotyledon takes one root-xylem. Duchartre (’48) mentions that 
R. montana and R. graveolens are always dicotylous. The only two seedlings of R. graveolens 
I have seen were both dicotyls with practically equal seed-leaves. 
2 I am permitted by Dr. Ethel de Frame to include here the results of an examination of a great 
number of Acer seedlings growing wild or in gardens. In all 7,190 seedlings were collected, and 
among these 143 showed some abnormality : this being 1-98 per cent, of the whole. The numbers 
of the different classes were as follows : 2 complete syncotyls, 2 hemisyncotyls, 44 hemitricotyls, 
72 tricotyls, 18 hemitetracotyls with two split seed-leaves, 2 hemitetracotyls with one split and two 
entire seed-leaves, 3 tetracotyls. A higher percentage of schizocotyls occurred among the offspring 
of cultivated than among that of wild trees. Schizocotylous seedlings of Acer pseudo-platanus have 
also been figured and described from the morphological point of view by Thiselton-Dyer (’02), who 
draws attention to the importance of an anatomical study. 
