444 
DEMETRIUS ION ANDRONESCU 
found that the embryo of MirabiHs germinates easily without its endosperm, 
although the later is necessary in the advanced stage of germination for 
the development of the plumule. He demonstrated that* the endosperm 
can be replaced by a paste made of triturated endosperm of the same seed 
or of seeds of different species, or by merely a paste made of starch to which 
have been added mineral nitrates and phosphates. He claimed therefore 
that the nutritive power of the endosperm does not depend upon its cellular 
organization. In another series of experiments, in which were involved 
Mirahilis longiflora, Canna aurantiaca, Aucuba japonica, and Phoenix 
dactylifera, the same author (14) concluded that ''The endosperm of oleagi- 
nous and aleuronic seeds has a probe activity; it digests itself. . . . The 
starchy and cellulosic endosperms are, on the contrary, passive; they are 
digested by the embryos. ..." 
Bolciszwesky in 1876, as quoted by Brown and Morris (3), experi- 
menting with embryos of oats, rye, and lupine which he germinated in 
moist soil, also concluded that the endosperm is not necessary to germina- 
tion. 
Brown and Morris (3) raised the question: "Is there any residual life 
in the cells of the endosperm.^" By applying, or, to use the authors' own 
term, by "grafting" new living embryos to seeds of barley which had been 
killed by the action of heat, chloroform, and alcohol, these embryos de- 
veloped almost normally, using these dead endosperms. These observa- 
tions supplement the statement of Sachs that "The endosperm of the 
grasses is a dead storehouse of reserve material ; therefore during the germi- 
nating period the life of the embryo is not parasitic but saprophytic." 
Stingl (12) found that the endosperms of the seeds of wheat, rye, barley, 
aiid o^ats in mMy cases may be substituted one for another during the 
process of germination. 
Dubard and Urbain (5) germinated embryos of seeds of barley, oats, 
and maize, without endosperm and cotyledon, in Knop's 0.5 percent 
nutrient solution. Some of their conclusions were: that the endosperm is 
not indispensable for the development of plants; that the endosperm is 
favorable in the first stage of germination; that the embryo possesses in 
itself the reserve material for its development; and that by mutilating the 
embryo in many cases a long vegetative development is obtained after 
which the plumule withers away, while the rootlets are very little advanced. 
From the work of the above named authors the following conclusions 
may be drawn: (i) The endosperms of grasses are dead storehouses of re- 
serve material at the disposal of the embryos; (2) That the endosperms are 
not essential for germination of the embryos. 
We must bear in mind however that the authors quoted studied the 
germination of seeds and embryos without endosperms and cotyledons in 
different media and nutrient solutions, merely in their first stages of de- 
velopment, but that no one of them extended his study to the complete 
