366 BOTANICAL GAZETTE [JUNE 
compensatory regulation or correlative growth cannot be considered 
as an unimportant factor. 
II. Methods 
The seedlings and plants examined in my work were Helianthus 
annuus, Phaseolus multiflorus, Ricinus communis, Brassica oleracea, 
Coleus tricolor, Fuchsia speciosa, and Vinca major. ‘The stems com- 
pared were selected with reference to similarity in size and vigor. 
All tension and normal plants were grown under similar conditions. 
When collected, the stems were numbered, cut into suitable 
lengths, and put into 50 per cent. alcohol for further examination. 
To study the stems, freehand cross-sections were made and the tissues 
measured by several methods. Further description of the different 
methods will be given under the proper headings. 
III. Results of experimentation 
A. INFLUENCE OF TENSION ON STEMS 
A number of seedlings were subjected to tension in the follow- 
ing manner: Strong, light twine was fastened about the stem, which 
had previously been wound with cotton flannel to eliminate any 
injury due to the cord. The twine was then run over a lightly rolling 
pulley, hanging from a support directly above, and the weights 
attached to the free end of the cord. The twine was so fastened 
that very little injury, if any, was caused. Two loops were made 
from short pieces of twine, each about 30°™ in length. These were 
noosed about the stem, from opposite sides one within the other, so 
that four circles of twine surrounded the stem over the cloth. The 
ends of the loops were then connected by a short piece of twine. To 
this was fastened, in an adjustable manner, the cord running over the 
pulley. In this way, if one loop should happen to be longer than 
the other, an adjustment took place, so that the strain on each loop 
was the same, in consequence of which the stem felt the pull in the 
direction of its longitudinal axis. 
When the seedlings were young, the unfolding leaves and growing 
tips might be somewhat hindered in their growth by the interference 
of the loops of twine. To prevent this, a strip of light whitewood, 
about 1o°™ long, 3™™ wide, and not more than 1.5™™ thick, was so 
placed between the loops as to form a diamond-shaped figure, such as 
