147 
Primitive A agio sperms. 
possess more than two cotyledons have little claim to be regarded as 
primitive on other grounds. Moreover, there is some reason to think that 
where more than two cotyledons are present they have been formed by the 
splitting of an original pair (Hill and de Fraine, 42 ; p. 473). 
Comparison with Gymnosperms is thus — so far as it goes — in favour 
of the assumption that two cotyledons were transmitted to Primitive Angio- 
sperms by their ancestors. 
Embryology within the Embryo-sac. 
The embryological evidence is divided, as already explained, under 
two heads. We will begin with the development of the embryo within the 
embryo-sac. 
The development of Capsella among Dicotyledons, and of Alisma 
among Monocotyledons, is described and figured in all textbooks. The 
figures are commonly reproduced from Hanstein’s memoir of 1870 (35). 
Hanstein himself described several variants on the course of events which 
he considered typical, and later research has brought others to light. The 
great majority of Dicotyledons hitherto investigated do not, however, 
differ much from Capsella. More variation seems to occur among Mono- 
cotyledons. 
In Alisma and Capsella the cotyledons are the first permanent organs 
formed in the pro-embryo. The growing-points of root and stem appear 
later. This is the regular order of development in Monocotyledons and 
Dicotyledons alike. Leaving such abnormal forms as Cuscuta (Koch, 51) 
out of the question, the only exceptions with which I am acquainted are 
found among the Papaveraceae. In Roemeria refracta , figured by Hanstein 
(35, PI. VII, P"igs. 6 and 8), the growing-point of the stem is indicated 
almost as soon as the rudiments of the cotyledons. Hegelmaier has shown 
that in Hypecoum procumbens the stem-apex appears at the same time as 
the cotyledons, or even a little before them (37, PI. Ill, P"igs. 23-6). 
The free end of the embryo increases in size in both types. In Capsella 
it bifurcates to form the two cotyledons : in Alisma the whole is trans- 
formed into a single cotyledon. The growing-point of the root appears at 
the other end of the embryo, separating it from the suspensor. It is alike 
in both types. The growing-point of the stem appears between the 
cotyledons in Capsella , but at one side of the pro-embryo in Alisma — just 
below the single cotyledon. 
To avoid error in the interpretation of these results, the goal must be 
constantly kept in view. The question before us is whether the Primitive 
Angiosperms possessed one cotyledon or two. Can the facts just rehearsed 
be used as evidence on either side ? 
The simplest interpretation is certainly that the single rudiment found 
in both classes remains undivided in Monocotyledons, but divides in 
