№ 
208 STUDIES IN PLANT REGENERATION 
region had been removed in the manner just described. These 
pieces were placed in a horizontal position in sand, with the straight 
edge down. In seven of the nine pieces thus placed, roots devel- 
oped as in the erect pieces, from the cut end of the basal inter- 
node. Тһе other two produced several small roots from the nodes 
on the straight side. Опе of the seven in which the cut end of the 
internode bore the roots, produced these as well from an oblique 
cut on the notched side of the stem, which was in the air, but not 
far above the ground. When eight pieces like those just men- 
tioned were placed with the notched edge under the sand, in all 
cases, again, the basal internode developed roots, and in two pieces 
the upper oblique cut produced them as well. 
These experiments prove, it seems to me, that the organ-form- 
ing “tendency,” as it may be called for the present, in this plant 
does not extend very far in from the edge of the stem. When, 
by means of the deeper cuts, all the region which is ordinarily 
active in this formation is removed, the roots are formed, not at 
the nodes, but only at the basal cut. Оп the other hand, when 
less has been lost, or when the oblique cuts have not excised all 
of this region, the organs may be formed as well or exclusively at 
these points. Moreover the inability of the parts to form shoots 
after they have been removed indicates the very narrow limits 
within which this regeneration takes place. 
Experiment 17.— On November 4, 1904, on one piece of 
Muhlenbeckia stem, the diagonal cuts were made along one edge, 
removing the buds, while the other edge was left uninjured. The 
whole side which had one bud present was placed in the sand. 
Roots developed from one of the lower uninjured nodes, and from 
the basal internode. Тһе former were removed and did not grow 
out again. А shoot, however, developed from the axillary bud. 
On October 15, 1905, the piece was taken out and examined. It 
was found that the shoot had established a direct connection with 
these basal roots by means of a secondary growth of vascular tissue 
through the middle of the old stem (Fig. 7). On section, it was 
found that from interfascicular cambia new bundles had been 
formed between the old in this restricted portion. The number of 
bundles intercalated between the old ones was most often two, but 
in places, опе or three ; and these bundles ran continuously from 
