466 STEVENS: ENDOSPERM IN VACCINIUM CORYMBOSA 
Wareham, Massachusetts, during June, 1916. The material was 
fixed in a solution of equal parts glacial acetic acid and absolute 
alcohol, imbedded in paraffin, cut and stained in the usual way. 
From a study of this material it is apparent that the development 
of the endosperm of this species may begin in two quite different 
ways, either by the formation of a cross wall following the first 
division of the primary endosperm nucleus as has been reported 
in the other genera of the Ericaceae, or by a period of free nuclear 
division as has been described in a very large number of species. 
Fic. 1 shows a typical two-chambered embryo sac much like 
those described in Monotropa and Epigaea. The material ex- 
amined also showed the four-chambered stage which characteris- 
tically follows this. Fic. 2, on the other hand, illustrates an 
embryo sac which has developed by free nuclear division. The 
stage represented by Fic. 3 may obviously have resulted from 
free nuclear division followed by the beginning of wall formation 
near the center of the embryo sac, or there may have been some 
free nuclear division after the formation of the cross walls. 
Fic. 4 shows a somewhat more advanced stage, in which the 
irregular arrangement of the cell walls suggests that there was a 
period of free nuclear division. In this figure the layer of small 
cells with dense protoplasm (shaded) represents the ‘tapetum” 
mentioned in Epigaea (5, f. 4). The developing haustoria may 
also be noted. The antipodal haustorium consists of only one 
cell, while the micropylar haustorium contains three. In the 
mature seed of Vaccinium corymbosum the haustoria are larger 
than in the seed of Epigaea repens; their development also begins 
relatively early. In Epigaea when the embryo is in the eight- 
celled stage and the endosperm well differentiated both haustoria 
are still small and consist of but a single cell, whereas in Vaccinium — 
the haustoria have attained a considerable size before any divisions 
of the fertilized egg are apparent. 
Hofmeister found that Vaccinium (3, p. 141) differed from the 
other genera described by him in that after the formation of the 
first wall across the embryo sac, endosperm developed only in the 
antipodal chamber, while in Monotropa and Pyrola endosperm 
developed in both chambers (see also Coulter and Chamberlain, _ 
I,p. 177). The condition found by the writer in V. —— 



