1905) MCCALLUM—REGENERATION IN PLANTS 119 
the inside of the tube. The air in the tube under these circumstances 
must have been saturated with moisture, all of course at the expense 
of the water in the stem. Within a week three of the stems and a few 
days later a fourth showed vigorous roots coming out, which grew 
rapidly and soon were several centimeters long. Here again the 
stimulus certainly was not any in- 
creased water in the stems. It seems 
as though a moist atmosphere outside 
of the stem can act as a stimulus with- 
out any increase, in fact even a de- 
crease, inside. How this could act 
through the epidermal and outer corky 
layers is not clear. At first it seemed 
that the real cause lay in the removal 
of the piece from all influence of the 
roots below, but glass tubes similarly 
placed around portions of longer pieces 
whose roots were intact and active 
below resulted in the production of 
roots just the same. Similar pieces covered with a thin coat of wax 
to prevent any evaporation showed no signs of root development. 
In submerged aquatics, where there is no current of water through 
the plant, but where the absence of a cutinized epidermis allows 
free diffusion in and out at every point, and where all the cells are 
constantly saturated, the removal of a part of the stem does not 
cause any change in the amount of water present. If such plants 
regenerate it is not due to disturbances in the water content. 
Experiment 37.—Portions of the stems of the extreme aquatic 
forms of Proserpinaca palustris and Ranunculus multifidus were 
severed from the parent plants and left submerged. In all cases 
New roots were at once organized and grew rapidly at or near the basal 
end, and at the other end shoots started from the latent buds in the 
leaf axils (fig. 14). Isolated pieces of roots of Taraxacum, Rumex 
crispus, and stems of Zamia all organized new shoots while they 
were still losing moisture. In cases like Salix there is no doubt that 
contact with water will start the development of roots along the stem, 
yet these can also be started by other causes while the cells are losing 
Fic. 14 
