582 Burns . — Heterophylly in Proserpinaca palustris , L . 
germination, the fourth was allowed to freeze and remain frozen during 
the winter. 
In March of the following year seedlings began to develop in all four 
lots. Those in ‘ a ’ were transplanted to two-inch pots and kept in as 
saturated an atmosphere as it was possible to obtain. Lot ‘ d ’ was 
removed to the greenhouse. We have now seedlings growing in water, in 
air — under the same conditions as Geraniums-— and in a saturated atmo- 
sphere. In every case, after the entire cotyledons, there appeared the 
‘ water-type ’ of leaf. I cannot say how long this would continue, as my 
seedlings died when they had from eight to twelve leaves. 
These experiments show that the young plant will produce a divided 
leaf regardless of external conditions as such. 
Cuttings. 
Reference has been made to the work of Familler on Campanula 
rotundifolia. Cuttings made from stems of this plant, which were producing 
linear leaves, reverted to the round form of leaf. To test this on Proser- 
pinaca a large number of cuttings were made from that part of the plant 
growing in air and producing entire leaves. These cuttings were made on 
May 28, rooted in sand, and finally transplanted to two-inch pots. They 
were kept in the greenhouse in comparatively dry air. They all grew 
luxuriantly, producing entire leaves on the main stem, blossomed, and 
ripened fruit (Fig. 1). It would thus seem that this plant differed from 
Campanula rotundifolia. However, on August 12, these plants were 
divided into three groups. One was left in the same condition as before 
and served as a control, another set was placed in an aquarium under water, 
and cuttings were made from the third which were planted in earth in air. 
The control plants continued to produce entire leaves. Those in the 
aquarium did not all act alike. In some the vegetation-point of the main 
stem stopped producing leaves and side-branches developed, always with 
the ‘water-type’ of leaf. In other cases the plant on its main stem 
reverted to the production of the ‘ water-type ’ of leaf. But most interesting 
was the behaviour of the cuttings. These were left under exactly the same 
conditions of light, heat, &c., as the control plants, and yet they began 
immediately to produce the ‘ water-type ’ of leaf. Thus there is a stage in 
the development of Proserpinaca during which a reversion may be caused 
by propagating by cuttings. 
A second set of cuttings was made from plants growing in water and 
producing the ‘ water- type ’ of leaf. These were started exactly as those 
in the first set. These cuttings were made in September. After trans- 
planting to two-inch pots they were divided into three sets. One set was 
placed in an aquarium under water; another set was placed in 5 cm. of 
water, thus immersing the root-system only; the third set was placed 
