254 BOTANICAL GAZETTE [APRIL 
TABLE VI. 
(Plus sign indicates g th into gelati taining air.) 
. No. of dae Growth 
Species fesd¥¥val bbe: Gases i Stamey! ae Btls Remarks 
Lupinus albus..| 1s 24 co, Air § 317 Gela ee thoroughly sat- 
urated with 
S 18 30 CO, Ar 2 Pat is Gelatin not thoroughly sat- 
‘upinus albus 16 24 co, Air 6 713 urated with CO,. 
Lupinus albu 13 30 mae Air = 7/1 
W s 7 30 Cc Air 7 20 
€ Ir 30 co, Air 316 { 8-28 prea retarded during first 
12-27 
Pisum sativum , 5 48 Air 4-20 
Lupinus albus.. ‘ éo H ee I 8 } St Growth much retarded. 
A frame (jig. 5) was made of strips of glass cemented together 
along the edges and fastened at the ‘ends in grooves in wooden 
bars (a and e). These strips were of two sizes. The wider 
ones (6) were so secured in the wood that narrow spaces were 
left between them (c), which were covered by the narrower 
strips (d), except in the lower 1.5%. Thus a series of little 
chambers was formed not more than 3™™ in diameter, open at 
the top and closed at the bottom except for the opening (/) 
between the strips of glass. By temporarily closing these open- 
ings with strips of glass the chambers could be filled with gela- 
tin, after the hardening of which the two strips of glass could 
be removed. This frame with the little chambers filled with 
gelatin was now pressed down into a battery jar containing 
enough water to cover the lower bars of wood (“). By insert- 
ing rubber tubing between the edges of the frame and the sides 
of the battery jar two almost air-tight chambers were formed. 
Down the middle of each column of gelatin a needle was 
thrust to make easy the passage of roots which were pushed 
down into the gelatin till the tips came below the glass covers, 
as shown in fig. 5. The chambers on the two sides of the parti- 
tion were then filled with gas as desired. 
In one series of experiments, in which CO, was passed slowly 
into one chamber and air more rapidly ee the other, of 3 
lupine, § pea, and 4 popcorn seedlings used, all grew nearly 
straight, though 2 took a somewhat slanting course toward the 
air chamber. 
In another series, in order to reduce the surface for the cross- 
