I HAVE briefly described in vol. xv. of the Linnaean Society's Journal, the nectar-glands found at the base of the fronds of the brake fern (Pteris aquilina) which are visited by ants for the sake of their sweet secretion. This case seemed to me to show in a striking manner that extra-floral nectar-glands are not necessarily protective in function, because the fern has, in England at least, extremely few enemies. The following extract of a letter lately received from Fritz Muller (of St. Catharina, Brazil) is of considerable interest in relation to this subject. He states that “the honey-glands on our Pteris aquilina serve, without doubt, to protect the ferns from the depredations of the leaf-cutting ants (Œcodoma), as is the case with Passiflora, Luffa, and many other plants. The glands of the Pteris are eagerly visited by a small black ant, Crematogaster, of which the Œcodoma seems to stand in great dread. On the other hand, when no protecting ants are present, I have seen Œcodoma gnawing the young fronds; here, as in other cases, it is only the young leaves that stand in need of protection, the older ones not being attacked by the leaf-cutting ants.” This fact might, no doubt, be used as an argument by those who believe that all nectar-glands were originally developed as protective organs, and this argument would have great force if it could be shown that Pteris aquilina is a form which has arisen in countries where protection is needed; but even in that case there would remain the difficulty of accounting for the continued functional activity of the glands in districts where no such protection is required. Or it may be said that in past ages the glands on our European Pteris served as a protection against enemies which have now become extinct. But here we are again met by the difficulty of accounting for the continued activity of the glands. It is characteristic of evolution that great changes occur in the functions of organs, and I think that it will generally be allowed that even the most beautifully adapted apparatus must have originated in an organ performing some comparatively simple function. The question at issue may perhaps be stated as follows:—In the cases where the nectar-glands are now well developed has there been a special course of structural development in close relation with the need of the plant for protection? Has there been a course of evolution such as we may believe has taken place in the formation of the food-bodies in Acacia spaerocephala and Cecropia peltata, or should we not rather believe that the sweet secretion has been developed in connection with some unknown process of nutrition; according to this view, a well developed system of glands may continue merely performing some obscure excretory function, and consequently, although the presence of nectar-glands has undoubtedly been of the utmost importance in determining the survival of certain species, yet it is hardly fair to assume that all nectar glands were originally protective in function. For many plants secrete large quantities of sweet fluid, which serves no such purpose. This argument is given by my father in his “Effects of Cross and Self-Fertilisation” (p. 402). In addition to the facts there given in support of this view, a curious case described by Prof. H. Hoffmann may be mentioned (“Ueber Honigthau,” 1876). He states that numerous large drops of sweetish fluid appeared on the under-surface of the young leaves of a camellia. He also alludes to a similar abnormal production of honey-dew on an ivy plant.