1913] MARTIN—POLLEN OF TRIFOLIUM PRATENSE 121 
tion may be a certain ratio between the amount of water taken up 
and transpiration. 
An effort was made to reduce this required water supply to 
some definite expression by running tests on bladder suspended 
over different concentrations of H,SO,  Gram-molecular solu- 
tions were placed in large, wide-mouthed bottles fitted with rubber 
corks. A glass tube about ro inches in length was run through the 
rubber cork and the membrane suspended from a cork fitted over 
the lower end of the glass tube. With the upper end of the tube 
corked, the apparatus was left in the required temperature 48 
hours to secure moisture equilibrium between liquid and membrane. 
If stored longer than 48 hours fungi gave trouble. The pollen was 
collected on the end of a glass rod and deposited on the membrane 
by running the rod through the glass tube. This method prevented 
interchange between outside and inclosed air. The percentages 
of moisture were approximated from data given in LANDoLT- 
*"- 7FIN (19). The humidity at the pressure of saturation over 
ater was considered 100 per cent, and moisture for each 
temperature used and the percentages over the solutions are based 
on the roo per cent. The results of two series run are given in full 
to show variation and the others are summarized (tables V-VII). 


TABLE V 
TEMPERATURE 35° C. 
i 5 f P t. of ie 
Gram neo of — a of odlien” po “ae — : = _— Condition 
Questa reser desk es 95.5 1201 0.88 Turgid 
2 gi eect Ae ae 96.5 2187 17.4 Some bursting 
ha eS 97.2 i) Ga 290.9 Little bursting 
OF ie eee cree 98.6 2209 10.0 20 per cent 
bursting 
Pure WO. Some less than 2736 24.4 30 per cent 
fete) bursting 







Mos ges os ke ex 95-5 1120 ° Turgid 
i ee eee 96.5 1396 0.21 : 
se ase mac rea 97.2 1590 12 S 5 
sc Se REED Ee Oe 98.6 1340 1.34 Little bursting 

