118 BOTANICAL GAZETTE [AvGusr 
In less than a week roots had appeared on all three near the base, 
The weight now was found to be 26%, a loss of 1.5%", most of which 
certainly was water. Another piece at first weighed 10.58; and 
when the roots coming out on it were 1°™ long the weight was 10.1%, 
Thus, with the piece as a whole losing water, but with a moist atmos 
phere outside, roots develop. : 
Experiment 34.—Two pieces of stem 30° long 
were fastened with their basal ends connected with 
the water faucet, so that the water was forced into 
.them under high pressure. This pressure was 
sufficient to cause water to ooze slowly from the 
opposite end of the stem. The air surrounding the 
stem was that of the laboratory. No roots any 
where enlarged enough to break through the cortet 
to the surface. Here a greatly increased amoullt 
of water in the stem does not start the roots. . 
Experiment 35.—Stems growing with vigorols 
roots in water cultures were selected. A fewin 
above the water a ring of bark 5™™ wide, cutting 
into the wood, was removed. Just above this for 
about 1°™ the bark (i. e., all outside of the wood) 
was plainly drier than at any other point, yet Ce 
root primordia in this part in a few days enla ge 
and broke through the cortex and epidermis, gil 
went no farther; but when one of the pieces ™# : 
put in moist air they grew out rapidly. 
Experiment 36.—Five pieces of stem of Salt 
jragilis, each 30°" long and one and two YP" 
old, were cut off from larger branches gf™ 
with roots in water. Two of the pieces bore young leaves, while 
on the other three the buds had not yet opened. Each epi 
carefully freed from any water on the surface, was placed inside & 
dry glass tube as small in diameter as would admit the piece (fg. 13 , 
Each tube was sealed air tight at each end by a rubber stoppe!® 
wax. Here there was no possibility of any absorption of mor 
on the other hand evaporation was constantly going on from the s 
face of the stems, and the moisture condensed in little droplets 
Fic. 13 
