OBSERVATIONS ON THE DEVELOPMENT AND GERMINA- 
TION OF THE SEED IN CERTAIN POLYGONACEAE 
Edward F. Woodcock 
The Polygonaceae have been included by Harz (7, p. 1072) under 
the ''Curvembryonaten" which he speaks of as being furnished with 
an abundant perisperm, and laterally placed embryo. The term 
"perisperm" is applied to the remaining portion of the nucellus in a 
mature seed. Harz (p. 1102) carefully figures and describes the 
buckwheat seed, considering the entire storage region as perisperm, 
and stating that the same relation holds in species of Rumex. 
Johnson (11, p. 334), evidently influenced by the work of Harz, 
was led to believe that there is a rather close relationship between 
the Polygonaceae and the Piperaceae, in which family he observed 
perisperm. In these Piperaceae he looks upon the very abundant 
perispenm as the real storage region, it being separated from the embryo 
by a layer of endosperm, which, instead of acting as a storage region, 
serves to digest and pass on food material to the embryo from the 
perisperm. In an earlier paper (10, p. 368) Johnson had already 
pointed out this restriction of the endosperm in Saururus and suggests 
the probability of a similar relation existing in all seeds furnished with 
a large amount of perisperm as in Polygonaceae, Chenopodiaceae, 
Phytolaccaceae, Caryophyllaceae, etc. 
Kraemer (12, p. 262) in his text-book merely mentions the fact 
that the Polygonaceae possess a "mealy" endosperm, without going 
into a discussion of the seed development. He also figures in detail 
a transverse section of Fagopyrum esculentum, but states nothing 
further in regard to its morphology. 
Two years later Stevens (18, pp. 59-65) made a detailed study 
of the morphology of the seed of buckwheat to determine whether, as 
suggested earlier by Harz and Johnson, perisperm did really exist 
in the mature seed. After careful investigation he was convinced 
that no perisperm existed in this seed at maturity. The fact that 
Humphrey (9), working on the Scitaminales, found, in closely related 
genera, various amounts of endosperm, suggested to Stevens that 
454 
