204 STUDIES IN PLANT REGENERATION 
capable of forming organs, but to a different extent. While the 
cortex can produce both kinds of structures — опе as the result 
of regeneration and the other in the normal manner— the callus 
formed by the pith is able to originate roots only. This was the 
only case where roots formed from a callus in the parsnip. 
The foregoing results indicate a close relation between shoot- 
formation in the parsnip and the action of some external factor, 
probably water. Vóchting has indicated that moist sand affords 
a complex of conditions, of which the water, and the contact with 
a hard substance are two. How the latter could prove active 
here is inconceivable. It is only for lack of any other evident 
cause, however, that one is willing to ascribe the difference in the 
behavior of the two surfaces to the greater amount of moisture 
in the sand. The air in a cutting-frame, while not saturated, is 
never dry, and amply suffices for shoot-formation in other instances. 
There are no other grounds, moreover, for considering water in 
liquid form as a usual factor in the production of these or- 
gans. The experiments that naturally suggest themselves as 
more definite approaches to the solution of the problem — such аз 
growing the parts in a saturated atmosphere, or entirely below the 
sand, or in sterilized water, or covered on both sides by moist 
sphagnum, cotton, or filter paper— were tried without success. 
None of the parts so treated resisted decay long enough to regen- 
erate. Pieces have recently been set up from which rectangular 
portions have been removed, thus giving an upper and a lower 
surface under the sand, but sufficient time has not yet elapsed to 
yield results.* 
As an example of a woody root, Pelargonium radulum was 
chosen because of an accidental observation resulting from another 
experiment. A leaf-cutting had been made, which rooted freely and 
produced a shoot from the end of the petiole. The new shoot, 
in turn, formed a large root-system of its own. In removing the 
young plant from the sand, after it had been growing five months, 
its root was broken, part remaining in the frame. From this res- 
idual piece, three months later, a new shoot grew up, which turned 
out, upon investigation, to have arisen from the middle of the up- 
* Beyerinck (7. с. 65) also found tha 
| t the apical surface of this root seemed рге- 
disposed to form shoots. Не offers no ex 
planation, however. 
