SYMMETRY OF THE CENTROSPERM X. 279 
Erect Variety of Atriplex rosea. 
Observations were made upon four specimens (Erl, Er’, Er?, Ert) of the erect 
variety of this species, and upon one specimen (I 28) that appeared to be transitional 
between the erect and the prostrate form. Between the two extreme forms several 
differences as regards mode of growth are revealed. 
In the erect variety :— 
(1) The main axis, at least at first, grows out more strongly in comparison with its 
branches; this would be anticipated, sinceit grows erect and has to raise aloft and carry 
the branches. 
(2) The development of the branches at the base of the plant is much weaker. In the 
specimens examined no branches were emitted from the axils of the cotyledons ; those 
emitted from the axils of the lowest two foliage-leaves were weaker than the immediately 
succeeding pair; sometimes the lowest two foliage-leaves subtended buds, though there 
were already branches in the axils of the succeeding pairs of leaves. 
In the £ransitional form the branches at the base seemed likewise to be intermediate 
in their behaviour. Weak branches sprang from the axils of the cotyledons; a weak 
branch was emitted from the axil of the lowest foliage-leaf, and a stronger one from the 
other leaf of the first pair, but the branches emitted from the axils of the next pair of 
leaves were no stronger than the latter. 
(3) The internode-curves and the subcurves are less regular than in the case of most 
of the specimens of the prostrate variety. It is possible that this greater irregularity is 
due to greater variation in the surroundings on the sloping sand, but it is also conceiv- 
able that the erect form is more unstable either because it has not yet reached the 
prostrate stage or has recently acquired its erect habit. The erect habit is certainly not 
due to the direct action of the environment, as both varieties are to be found growing 
side by side on the sand immediately at the foot of sand-dunes; though it is quite 
possible that the greater length of the early internodes of the erect form may result 
from the greater shade of the habitat. 
(4) The internode-curves are clearly compounded of two subeurves. (The curves are 
not printed.) 
Atriplex rosea Er’. 
| | | | | Whole 
| | | | | | | | | length. 
Main stem ..|8 | 2 | 195 1 | 29| 55 85} 7 1885|20 225 175 12|3 6 |Flower| 230 
| | | | | | | | | e P: 
| [ Branch III .. 74 | | Exp T.: | ve | 19 
Soares te ae LA usd seen] | Pd | B 
| | | | | | | 
| | | | | | | | | | E 
Hos dV. | 16] 5 aes Ü ee i rici 35 
Hong ICCIUgy [190 0 340 DNA eae eq i get | | 5 
| | | | | | | | | | | 
| ow . .| Not measured. | | | | | | | | 
| D | | | | | | | | | | | | á 
ay ye . 182 LoT | 9 | 4 zr | 3 . | .> . | -s | ee | | 45 
| 
| 
na TA N tee ee 
