240 BOTANICAL GAZETTE [NOVEMBER 
for 3-4 days, the dry tip of the grain exterior to the tube was sec- 
tioned and treated directly with the Schryver formaldehyde reagent 
(29). With long periods of exposure to high concentrations of 
formaldehyde (1-8) penetration appears to be possible at either 
tip of the grain or on either face. The second method employed 
was to measure the degree of semipermeability of the seed coat 
indirectly by determining the weight increase of the seeds when 
soaked in distilled water and in formaldehyde respectively. For- 
Ses es em 
| | | 
, 
/ aie 
_ AT Het a ae ia 
el ae ae Sg) S| i 
an ez 
oe —— 5 
, ae oo f 
4 
4 
a 4 
4 , 
F) 
i 
t 
H 
t 
PENETRATION 
HOURS 6 ny 47 1 2 247 29: 
1G. 5—Comparative absorption by wheat of water and formaldehyde mixed 
1 to 8 parts of water, Hybrid 128 wheat from Moro, Oregon: solid lines indicate per- 
centages of weight increases when seeds were soaked in formaldehyde, broken lines 
hen soaked in water. 
maldehyde of high concentration (1-8) was used, in order to make 
more conspicuous any semipermeability differences of the coat 
toward the water and formaldehyde. In harmony with the results 
of Brown (15), if seeds with semipermeable coats be placed in salt 
solutions, water will be taken up by imbibition until the inwar 
force is offset by the equal outward osmotic force incident to the 
solution outside. If the two forces just balance each other, and 
if the coats be perfectly semipermeable, further soaking of the seeds 
in solution will not cause a rise in the curve indicating percentage 
weight increase, and the curve will continue horizontal. On the 
