344 BOTANICAL GAZETTE [way 
These cylinders or stockings were slipped over immersed 
roots as the seedlings were suspended over basins of water like 
those already described. When the basin was set in revolution, 
the stockings behaved well, but results were not forthcoming. 
The failure was probably due in part to the interference of the 
cylinders with the direction of the water stream, and in part to 
their mechanical interference with the curving of the roots. 
3. Experiments with glass tubes. 
The third and most satisfactory method for determining the 
extent of the sensitive area was the employment of glass tubes 
of several patterns, by means of which definite parts of the roots 
were exposed to the water streams in the revolving glass basins, 
and definite parts were shielded from the streams. The expeti- 
ments are divided into two groups, those testing the sensitive- 
ness of the apical millimeter or millimeter and a half, and those 
testing the rest of the root. 
a. Testing the sensitiveness of the apex of roots. 
To determine the sensitiveness of the apex of roots, straight 
glass tubes 3 to 4™ in length and 3™ internal diameter ee 
used. Seedlings of suitable size, usually not more than a half 
centimeter longer than the tubes, were selected for experiment 
The tubes were held in water and the seedlings passed in 
which a little wet cotton was pressed in around the seedling 
the use of a needle whose point was broken off. The i 
was pressed in firmly enough to hold the seedling im post" 
yet not too firmly, to allow the seedling to be moved up oF mo 
the tube. Preparations thus made were fastened to bars of ' ee 
_ by means of strips of blotting paper and rubber bands, # — 
a in fig. 75. The bar of wood, with its load, was next ise Ge : . 
_. with the glass tubes dipping into water in a glass — 
_ centrifuge, the seedlings were carefully adjusted meee ing 
. tips Projected 0.5 to 1™ from the glass tubes, and mee oe 
_ apparatus was set in motion. In recording the — .d 
below, data as to temperature and velocity will pew 
