278 BOTANICAL GAZETTE [OCTOBER 
The results from A. cernuum were nearly the same as for A. ¢ricoc- 
cum. Ninety-five embryo sacs were examined in the eight-celled stage. 
The egg apparatus was found in all of them, while antipodal cells were 
found in only twenty-nine; and, as in A. ¢ricoccum, these were invari- 
ably small, and apparently about to disappear. It was seldom that 
more than one or two could be found. Of the ninety-five specimens, 
thirty were collected on or after August 16, and no trace of antipodal 
cells could be found in any of these. Fifteen embryos were examined, 
all of which were normal in position and number. No antipodal 
cells were present in any sacs in which the embryo had begun to 
develop. 
My collection of A. Canadense was made from a patch covering 
about half an acre, at West Pullman, Illinois. In nearly every speci- 
men the nucellus had died long before the stage when fertilization 
might have taken place; and later in the season it was found that only 
six embryos had developed from the whole patch in which there had 
been thousands of blossoms. All of these embryos, however, were in 
the normal position. 
Since in A. fricoccum only about 21 per cent., and in A. cernuum 
about 30 per cent. of the sacs examined contained antipodal cells, and 
these cells in all cases were small and not found at all except in the 
earlier stages, the development of embryos by antipodal cells in these 
Species seems very doubtful CLARENCE J. E_tmore, Crete, Vebraska. 
