349 - BOTANICAL GAZETTE [MAY 
Now, suppose the cotyledons to become completely united 
by both margins, and that in the end this union is congenital. 
In order to develop freely, the stem bud must make its appear- 
ance at one side of the united members. As a rule, the coty- 
ledons are differentiated some time before the growing point of 
the stem appears; in this case the single rudiment which repre- 
sents both cotyledons is naturally formed in a straight line with 
the suspensor and the future axis. The rudimentary plumule 
when formed is forced to take up a lateral position (Alisma. 
jig. 3). As a rule, it attains the terminal position after germi- 
ALISMA CORY DALIS 
Fic. 3. Alisma Plantago (after HANSTEIN). Fics. 4-5. Corydalis cava (after 
SCHMID); fig. 4, front view; Jig. §, side view. Fic.6. Tamus communis (after SOLMS- 
LAUBACH). 
nation, forcing the cotyledonary member to one side. In Tams 
and Commelyna the growing point of the stem appears terminal, 
the cotyledonary rudiment lateral, from the first 5 (jig. ©). 
The single cotyledon in the ripe seed of Corydalts cava strik- 
ingly resembles that of Tamus in form and position (cf. jigs. 4 
and 5 with fig. 6). The simplest explanation of both structures 
is that they represent the fusion of two ancestral cotyledons, but 
that on account of the early formation of the rudimentary 
plumule, or perhaps because the cotyledons have united by one 
margin only—as in Ranunculus Ficaria—the stem bed has never 
lost its terminal position. 
This hypothesis accounts for the structure of the monocoty- 
ledo nous embryoina very simple way, and avoids the difficulty of 
*5SOLMS-LAUBACH, Ueber monocotyle Embryonen mit scheitelbiirtigem Vegeta- 
tionspunkt. Bot. Zeit. 36:65. 1878 
