170 BOTANICAL GAZETTE [SEPTEMBER 
SCHRODER (33), who found the same kind of selectively permeable 
coat in wheat. Arxrns (5) failed to find this character in the coats 
of beans. He found that the absorption of water from living and 
dead seeds was identical until germination commenced, at which 
time the osmotic phenomena of the living cells were manifested. 
The forces concerned in the initial stages of water intake are, 
according to ATKINS, those of capillarity and imbibition; but on 
germination, osmotic pressure begins to influence the amount of 
water taken up by the living seeds. As ScHRODER has pointed out 
(33, p. 188, footnote), ArKrns failed to take into account the open 
micropyle of the Leguminosae used; but the same seeds on moist 
sand, with the micropyle turned up, absorbed 90 per cent of 
their dry weight from 10 per cent NaCl in 6 days, according to 
SCHRODER. And since the present paper was written, TJEBBES 
(35) has found that the seeds of the sugar beet probably have a 
selectively semipermeable membrane as part of the inner seed coat. 
Up to the present, therefore, semipermeable membranes have 
been reported only in the Gramineae and Chenopodiaceae. How- 
ever, another interesting discovery was made in 1907 by BECQUEREL 
(7), who showed that the thoroughly dried seed coats of certain 
plants were impervious to various gases and to such penetrating 
substances as absolute alcohol, chloroform, and ether. He made no 
attempt to determine whether these coats were also semipermeable. 
During the last two years I have been investigating the char- 
acter of the seed coat of Xanthium with special reference to the 
work of BECQUEREL, Brown, and ScHRODER, and present here the 
results of the work. I wish to acknowledge with thanks the 
encouragement and helpful advice of Dr. WiLLIAM CROCKER, and 
to express my appreciation of the excellent facilities afforded me by 
the Hull Botanical Laboratory. 
II. Experimentation 
I. MOISTURE AND PERMEABILITY 
The discovery by BEcQuerEL that various gases, and alcohol, 
chloroform, and ether would not penetrate certain seed coats if 
completely dried, seemed so unusual that attempts were made to 
repeat his experiments, using the testa of Xanthium glabratum 

