586 Burns . — Heterophylly in Proserpinaca palustris , L. 
the other in water, were placed in the dark room. Within a few hours 
all the side-branches as well as the main stem were growing erect. The 
vegetation-point of stems growing in water soon appeared above the surface. 
This experiment was repeated with a large number of plants, and it was 
found that light was the determining factor. When light was removed 
the stems responded to the geotropic stimulus and turned up. When the 
plants were returned to the light the curves were straightened and the 
branches continued to grow horizontally. 
The answer to our question seems clear. At one stage of its develop- 
ment Proserpinaca is positively heliotropic and at another it is diahelio- 
tropic. That is, this plant changes its relation to light at different stages 
of its development. Furthermore, we find that the production of the 
‘water-type’ of leaf is intimately connected with that stage when it is 
diaheliotropic ; the ‘ land-type ’ of leaf with that stage when it is positively 
heliotropic. 
Finally, we find that the primitive form of leaf is always the ‘water- 
type * ; that side-branches developing in the air from a plant whose main 
stem is producing entire leaves develop the ‘ water-type 5 ; that all stems, 
regardless of all external conditions which I could control, produce the 
‘ water-type 5 in the fall ; that stems whose vegetation-points were removed 
in June threw out side-branches with the f water-type ’ when other plants 
under the same external conditions were forming the ‘ land-type.’ On the 
other hand, we find that at the time of flowering only entire leaves are 
formed ; that in summer almost every plant, whether in water or in air, 
produces the ‘ land-type ’ of leaf ; that the change from c water-’ to ‘ land- 
type 5 takes place earlier on strongly growing than on weak stems, and is 
dependent to a certain degree on external stimuli ; that the plant in its 
positive heliotropic stage forms the ‘land-type’ of leaf, and in its diahelio- 
tropic stage the f water-type.’ 
These results do not point to one or more definite factors which will 
explain the form of the leaf in Proserpinaca palustris. It is evident the 
water-environment is not the cause of the division of the leaf. Nor 
does it depend upon light, temperature, gaseous content of the water, or 
contact-stimulus as such. The only conclusion that seems justified by my 
experiments seems to be that Proserpinaca palustris has two forms — an 
adult form and a juvenile form (10). Under good vegetative conditions 
it has a tendency to produce the adult form with the entire leaf, blossom, 
and fruit ; under poor vegetative conditions it has a tendency to produce 
the juvenile form with the divided leaf. And furthermore, a reversion to 
the primitive form may be caused by unfavourably influencing the vegeta- 
tive conditions. 
