
120 BOTANICAL GAZETTE [AUGUST 
varied from almost zero to 96 on different membranes. Still, 
on membranes with the low percentage of germination, the per- 
centage of germination was high in the region where it occurred. 
In table IV are given the results obtained on a small piece of bladder 
that gave fairly uniform germination. 


TABLE IV 
Total number of pollen Number of germi- | Percentage of ger- 
grains nations mination 
COE IS a er ete 87 | 81+ 
1 EES a rs a a Ito | 81+ 
ee isesn 183 87+ 
1 jo a ee 170 96+ 
2 SS a ee he IOI 80+ 
DU ess 144 94+ 


The time required for germination at room temperature was 
8-10 minutes. This agrees with SANDSTEN’s (16) report on T. 
hybridum and T. repens, pe SOB TOF 
The lengths of tubes produced were various. The me. ~ 
length of tubes measured was about 15 times the diameter of the 
pollen grain. The lengths of the majority ranged from 6 to 15 
times the diameter of the pollen. It is probable that much longer 
tubes would have been produced if the water delivery of the mem- 
brane had remained constant. 
The results obtained with the membrane and parchment paper 
showed that the water supply was at least the important factor 
if not the only factor in determining germination. 
An attempt was made to secure the proper water supply by 
means of sugar solutions. Sucrose solutions with a difference of 
0.0877 volume-normal and ranging from 0.731 to 2.2 volume- 
normal were used. The only germination obtained in these solu- 
tions was less than o. 5 percent in 1 .7 volume-normal. In account- 
ing for this failure, three things should be considered: (1) the 
ree. of water supply permitting germination may be so small 
that it was missed by these concentrations; (2) the supply of 
rie tos and carbon dioxide might have been limiting factors since 
the higher concentrations were greater than those run under the 
increased pressure of these gases; (3) the condition for germina- 
