8o6 
Compton. — An Anatomical Study of 
certain species of Persoonia (Proteaceae), Nuytsia , and Loranthus (Loran- 
thaceae). 1 Other reputed examples have proved to be instances of deeply 
bifid cotyledons. The view that dicotyly is primitive in Gymnosperms 
and Angiosperms alike attaches much importance to the occurrence of 
bilobed seed-leaves ; thus Hill and de Fraine 2 consider that in Conifers ‘ the 
polycotyledonous condition has been derived from the dicotyledonous con- 
dition, in the vast majority of cases by the splitting of the seed-leaves and 
by the promotion of cotyledons from a lower to a higher rank \ Fletcher . 
also takes the same view for Persoonia and Nuytsia? The reverse view 
either takes no account of these abnormalities or regards them as stages in 
fusion rather than fission. 
The cases of teratological schizo- and polycotyly, which will be con- 
sidered below, are of interest from whichever point of view we regard the 
phylogenetic problem. If dicotyly be primitive they afford an instructive 
parallel to the course of evolution of polycotyly; if dicotyly be a derived 
condition these teratological seedlings may be regarded as reversions to 
a more primitive type. 4 
A few examples of polycotyly among normally dicotylous species have 
been studied anatomically, the investigators relying on chance specimens. 
I have been so fortunate as to receive from Professor H. de Vries a number 
of packets of seed of several of his schizocotylous races ; and although this 
seed was mostly harvested in 1894, the year when he discontinued his 
breeding experiments on these races, a large percentage of most of the 
species germinated and yielded abundant material, exhibiting all degrees of 
the abnormality. It is mainly on this material, for which, as well as for 
seed of Helianfhus annuus syncotyleus , my warm thanks are due to Professor 
de Vries, that the following statements are founded. 
I shall proceed to give some details of the structure of certain schizo- 
cotyls, following this by a discussion of the problems of polycotyly. 
Cannabis sativa (Urticaceae, or Moraceae). 
Dicotyls. The hypocotyl is from 6 to 9 cm. long and 0-9 mm. 
in diameter throughout: the cotyledons are oblong-ovate, 11 by 5 mm.; 
their venation varies considerably, but usually there is a well-marked 
midrib in the basal half which may or may not be prolonged to the apex 
(Fig. 20). 
The root has a continuous diarch xylem-plate flanked by two bands of 
phloem (Fig. 21). The xylem-plate begins to divide transversely at the 
collet and the phloems divide simultaneously : half-way up the hypocotyl 
the condition shown in Fig. 22 is attained. This slowly changes, triads 
being formed about three-quarters up the hypocotyl (Fig. 23), and each 
1 Duchartre (’48), von Miiller (’82), Fletcher (’09), Hill and de Fraine (’12). 
2 Hill and de Fraine (’09 ft ), p. 225. 3 Fletcher (’09), p. 878. 
4 Dorety (’10), p. 428. 
