278 MR. C. DARWIN ON THE ACTION OE 
and the chlorophyll-grains, though generally much swollen, were 
still distinct. It may therefore be concluded that in Sarracenia 
the chlorophyll-grains often undergo aggregation under the 
influence of carbonate of ammonia, but that they are less easily 
acted on than those of Dionea and Drosera. 
Leaves with Glandular Hairs and other Leaves.—1 had formerly 
observed, as described in my ‘Insectivorous Plants, that the 
glandular hairs of some plants absorb carbonate of ammonia and 
animal matter, and that aggregation is thus caused in them. 
Consequently such leaves and others without hairs were immersed 
in solutions of carbonate of ammonia (4 and 7 to 1000) gene- 
rally for 24 hours. No marked effect was produced on the chlo- 
rophyll-grains, excepting their occasional displacement, in the 
following cases (plants were selected almost by hazard, but 
which belong to different families) :—first, of leaves not bearing 
many or any glandular hairs, namely those of Brassica, Fumaria, 
Fuchsia, Robinia, Oxalis, Tropeolum, Euphorbia, Stapelia, Beta, 
Allium, Lemna, a fern (Nephrodium), a Marchantia, and a moss. 
Nor were the grains acted on in two species of Saxifraga (except 
on one occasion, when they formed masses shaped like a horse- 
shoe, presently to be described), nor in Primula sinensis—although 
the leaves of these three species are clothed with glandular hairs, 
which absorb carbonate of ammonia and undergo aggregation. 
Young leaves of Dipsacus sylvestris were immersed for 24 hours 
in a solution of 7 to 1000, and large yellowish highly refracting 
spheres were formed in the upper epidermic cells which do not 
include any chlorophyll-grains, and the grains were not at all 
aggregated in other parts of the leaf. "When the sections were 
irrigated with acetie acid or with alcohol, the spheres in the 
epidermal cells disappeared quickly, in nearly the same manner 
as occurs with recently aggregated masses in the cells of 
Drosera. 
Leaves of Cyclamen persicum, which bear hardly any glandular 
hairs, were left in a solution of 7 to 1000 for 43 hours, and this 
caused the chlorophyll-grains to collect into heaps ; in some parts 
the grains retained their outlines distinct; but in other parts 
they formed perfectly homogeneous bright-green masses of the 
shape of a horseshoe. These were cleared by alcohol; and it 
was evident that the grains had become completely fused toge- 
ther. Itis remarkable that many of the central cells near the 
vascular bundles contained spherical or oddly shaped confluent 
