320 BOTANICAL GAZETTE [ NOVEMBER 
forms cell walls. Some of the free cells accumulate in the antip- 
odal region and form a peculiar thallus-like mass, which further 
assists in obscuring whatever remains of the antipodals might 
otherwise be found (figs. 35-39). This growth was for some 
time a puzzle as to its origin, whether the result of division of 
the antipodals, or a growth of endosperm arising from the first 
division of the definitive nucleus and cut off by a cell wall from 
other endosperm, as is the case in Sagittaria.? All doubt was 
dispelled when the remains of the antipodals were found (gs. 
35,36) while this thallus-like growth was in early stages of 
development. In many cases the appearance was so peculiar 
and the connection with all other endosperm so obscured, while 
the entire mass stained so similarly to the embryo, that it might 
easily deceive one as to its nature. Its function is not clearly 
evident, unless it may be considered as a reserve of food material , 
after the suspensor ceases to supply nourishment. 
COMPARISON WITH OTHER DICOTYLS. 
Chamberlain? says the archesporial cell in Salix divides into 
a tapetal cell which sometimes gives rise to a tier of five or six 
cells but occasionally does not divide; and a macrospore mother 
cell which may or may not divide. If it does’there is a potential 
macrospore which sometimes divides and a vital macrospore 
which develops without further preliminary division into the 
embryo sac, 
In Aster Novae Angliae+ he reports that after the expected 
division resulting in four cells, the lowest usually develops into 
the embryo sac. 
Coulter’ says of Ranunculus multifidus: “In no case was @ 
primary tapetal cell cut off, the archesporial cell dividing directly 
into mother cells.” Alyssum might therefore be considered 
more primitive in this respect than any of these except Salix, 
which shows a very unsettled state of affairs. 
In the development of the embryo of Alyssum there is much 
? Schaffner, Bor. Gaz. 23: 252-273. 1897. 4Bor. Gaz. 20: 205-212. 1895- 
3 Bor. Gaz. 23: 147-179. 1897. 5 Bot. Gaz. 25: 73-88. 1898. 
ee 
= 
