The Dispersal of Seeds by Ants. 89 
had been detached. But whether the seeds of Ulex are brought 
into the nest or not, it is obvious from the experiments detailed 
above, that Formica rufa can and does carry the seeds about. 
We may therefore, I think, place the gorse and the broom in 
the category of plants, the seeds of which are dispersed by ants in 
addition to the earlier dispersal within a nearer radius, by the 
explosive contraction of the ripe seed pod (autochory). 
The reasons for including Ulex and Sarothamnus among the 
myrmecochorous plants are the following :— 
(1) . The seeds are provided with a bright coloured caruncle 
containing oily food-material and resembling in structure and 
contents the elaiosomes of other myrmecochorous plants like 
Chelidonium (the Greater Celandine). 
(2) . Ants are particularly attracted by the oil-containing 
caruncle, and as the experiments have shown, can and will carry 
the seeds of the gorse about. 
(3) . The curious rectilinear distribution of gorse-bushes 
along actual or disused paths and roadways as illustrated by the 
instances on the Yorkshire Moor 1 and the Derbyshire plateau does 
not seem explicable by ordinary dispersal due to the explosive 
dehiscence of the dry capsule, but can be paralleled by the 
distribution of such plants as the Celandine along the ant-runs. 2 
Whether or not some other interesting peculiarities in the 
distribution of the gorse depend on the presence or absence of 
certain species of ants in certain localities remains still to be shown. 
Geze has commented 3 on the absence of the gorse from the schists 
in the neighbourhood of Villefranche, while it seems to grow freely 
on the adjoining granites and clays, and this though topographically 
and climatically the same conditions obtain. Geze is inclined to 
attribute this selective difference in the distribution to the fact that 
the schists on weathering give a very poor soil, poor even in 
potassium, while the granites yield a soil poor in lime and phosphoric 
acid, but rich in potassium. The French observer is still investi¬ 
gating this problem in the remarkably local distribution of the gorse, 
mainly from the chemical point of view ; but possibly some attention 
to the possible predilection of ants for one or other of the two 
geological formations might help to settle this interesting question. 
1 New Phytologist, Vol. VII., No. 1, 1908. 
* See Kerner and Oliver. Nat. Hist, of Plants, Vol. II,, p. 867. 
1 Geze, J. B. Notes d’edaphisme chimique—Distribution de 
l’Ajonc (Ulex Europseus), aux environs de Villefranche-de- 
Rouergue. Bull, de la Soc. bot. de France. Tome 55, No. 6, 
1908, p. 466. 
