Primitive Angio sperms. 155 
fectly understood. But comparison of the normal type with certain ex- 
ceptional forms is very suggestive. 
Solms-Laubach (82) has described the development of the embryo 
in certain genera belonging to the Commelinaceae and Dioscoreaceae. He 
has shown that in such forms the cotyledon is lateral from the first, while 
the stem-bud is terminal. In his figures of Tamils communis the develop- 
ment of the cotyledon is traced from the beginning. It appears as a circular 
ridge surrounding the region in which the stem-bud will afterwards be 
formed. Growth in one part of the ridge soon ceases ; the opposite region 
grows rapidly, and the rudimentary cotyledon finally arches over the' hollow 
which contains the stem-bud. In the end the mature embryo is much 
shorter and broader than that of Alisma , but does not differ very widely 
from it in general shape. 
The embryo-sac is not included in these drawings, but figures in 
Le Maout and Decaisne’s textbook (52, p. 795 ) show the berry and ripe 
seed whole and in section, as well as the seedling during germination. The 
ovules are anatropous ; the embryo-sac therefore is not bent upon itself 
as in Alisma and Capsella. The seed is very little longer than broad, and 
the embryo has in consequence plenty of room to develop laterally during 
the process of maturation. In Commelinaceae the ovules are peltate (52, 
p. 8 68 ), and of course much broader than long. Here again the embryo has 
had ample elbow-room within the embryo-sac. 
This association of a lateral cotyledon with an unusually wide embryo- 
sac is very suggestive. Perhaps the apparently terminal position of the 
cotyledon in Alisma and the majority of Monocotyledons may have 
something to do with the long narrow embryo-sacs which are so common. 
The complete explanation would probably depend on other conditions too. 
We know far too little of the environment and needs of the growing embryo 
to attempt such an explanation yet. 
Three hypotheses to account for the terminal position of the cotyledon 
in Alisma and most Monocotyledons have now been discussed. First, that 
it is a terminal member and therefore no true leaf. Secondly, that it 
is a terminal member and one of the few leaves surviving to represent 
the original or ancestral leaf, which according to this view was essentially 
terminal. Thirdly, that it is a lateral member, forced into a terminal 
position by causes which we do not fully understand, but believe to be 
adequate to the task. 
Of these three suggestions the last seems the most satisfactory. If 
adopted, what bearing has it on the original question-^the number of 
cotyledons possessed by the Primitive Angiosperms ? 
Suppose the single cotyledon of the ancestral Monocotyledon was 
lateral, was it derived from a single lateral cotyledon in the primitive 
M 2 
