OF TENTACLES OF RORIDULA. 93 
Drosera is the base of the pedicel, and concludes that the 
lower part of the tentacle represents a prolongation of the leaf. 
Confirmation of this view is supplied by the terminal tentacle of 
Roridula, of which the basal part of the pedicel is a portion of 
the tissue of the leaf, while the upper part is of the nature of 
a hair. The other tentacles of Roridula are obviously hairs, 
but in some of the taller ones there seems to be a tendency for 
the tissue of the leaf to become merged in the base of the pedicel. 
in the case of Drosera, all the taller tentacles have undergone 
this modification, their basal part representing a prolongation of 
the tissue of the leaf, their upper part being of the nature of 
hairs, so that movement is limited to the basal part. 
I was unable to determine whether all the tentacles of 
Roridula are similar in function or not, but the evidence 
seems to show that the tall marginal tentacles are losing their 
digestive function, and becoming of the nature of catching 
tentacles, and we should have then a foreshadowing of the 
division of work amongst the tentacles that is characteristic of 
Drosophyllum, with its catching pedicellate tentacles and 
digestive sessilé glands, which are alike, structurally, in the 
possession of two layers of gland-cells—a middle layer, and a 
tracheid strand. Sessile glands are absent from the leaves of 
Roridula,; there we have smaller tentacles, only simpler in 
structure. I would suggest, on the basis of this comparison, that 
the sessile glands of Drosophyllum are derived from the pedicel- 
late rather than that the converse development has taken place. 
The primitive form of the droseraceous tentacle was evidently a 
hair, and length of pedicel and complexity in the structure of 
the head have increased equally. When these tentacles were 
able to attract and capture insects, the insects would ultimately 
sink down on the surface of the leaf, bearing the tentacles with 
them, because if these remained standing out from the leaf they 
would not be able to digest or absorb any nutritive matter ; 
this bending of the tentacles would be apt to break them. This 
is prevented, in the case of Drosera, by the power of movement 
at the base, which brings about incurving and subsequent 
re-erection of the tentacle. In Dyosophyllum this power of 
movement has not been acquired, and in order that the head of 
the tentacle may be brought near the insect the pedicel has 
