The development of the endosperm in Vaccinium corymbosum 
NeiLt E. STEVENS 
(WITH FOUR TEXT FIGURES) 
The occurrence of a “‘chambered” embryo sac, formed by the 
development of a transverse wall following the first division of the 
primary endosperm nucleus, has been reported in four genera of 
the Ericaceae. In recent studies of the high bush blueberry, Vacci- 
nium corymbosum L., the writer has noted a variation from this 
of sufficient interest to warrant brief publication. 
In 1849 Hofmeister (2) published an account of the develop- 
ment of Monotropa Hypopitys. According to his descriptions and 
figures (pl. r2, f. 11) the embryo sac in this species is first cut in 
two by a cross wall near the middle; afterwards cross walls are 
formed in each half and the micropylar end is cut off by a wall, thus 
resulting in an embryo sac of five superimposed cells. Nine 
years after the publication cited above Hofmeister described the 
development of Vaccinium Myrtillus and Pyrola rotundifolia (3), 
in both of which he found that the development of the endosperm 
took place as in the Monotropa without free nuclear division. 
Koch (4), in his studies of Monotropa Hypopitys, largely con- 
firms the work of Hofmeister. His figure (p/. ro, f. 12) closely 
resembles that of Hofmeister except that he finds no cross wall 
near the micropylar end and thus figures a four-chambered, in- 
stead of a five-chambered, embryo sac. 
In the trailing arbutus, Epigaea repens, the writer (5) found 
a condition much like that reported in other genera. His figure 
of the Epigaea (f. 3) very closely resembles Koch's figure of the 
Monotropa, a fact which the writer apparently overlooked in his 
earlier paper (5, p. 540). The writer also reported the occurrence 
in the Epigaea of haustoria extending out into the tissues of the 
integument from the ends of the developing endosperm. 
The slides of Vaccinium corymbosum, on which the present 
notes are based, were all made from material collected at East 
