was a habit which was acquired by degrees, but at what time it became the rule, would be now as difficult to determine, as whether originally there were one or two generations produced in the year. It is probable that at first there was only one generation developed annually, as is still the case with the purely agamous species. In any case, I consider it certain that parthenogenesis is the primitive mode, and that sexual reproduction is subordinate to it. It is of the greatest importance to the proper comprehension of the whole subject of alternating generations, to be able definitely to distinguish one generation as the primary one. The important differences existing between the generations of our time undoubtedly demand long periods for their development, while any certain means for estimating such periods are wanting. If gall-flies were found among fossil insects, we should have a definite starting-point, but no such discovery has been made. We only know that in earlier epochs a totally different climate to the present existed; under the powerful influences of a constantly though gradually changing climate this remarkable alternation of generations has been developed, while the adaptation to new conditions of life has left the general organization of the species more or less changed. A glance at these totally different generations as they now occur shows above all how difficult is the problem to seek 'in der Erscheinungen Flucht den ruhenden Pol.' By a few traits only, like hieroglyphics sculptured on the stones of antiquity, does alternation of generations demonstrate to us how remote is the period over which the history of its origin extends. SCHLESWIG, POSTSCRIPT. HOWEVER interesting may be the facts of alternation of generations as they occur in the oak gall-flies, I have no desire to ignore those species which live on other plants, or to omit the study of their evolution. The rose Cynipidae, of which I have observed three species, Rhodites rosae, eglanteriae, and spinosissimae, are always propagated by one annual generation. It is the same with the species of the genus Aulax, A. rhaeadis and A. hieracii; and with those of the genus Diastrophus, D. rubi and D. glechomae. On the other hand there is a Cynips on the maple which shows absolutely the same biological cycle as the Cynipidae of the oak. In this species, as in those, one agamous alternates with one sexual generation. The following is a description of this insect and the history of its development. 1. Pediaspis aceris, Förster-sorbi, Tischbein1. Gall. The gall is found on the roots of the sycamore (Acer pseudo-platanus), but it is also met with on the Norway maple (Acer platanoides). When solitary, the gall is round, like a pea, and about 5 mm. in diameter. When in clusters, these are found embracing rootlets, even those of 1 cm. in 1 Dr. Mayr of Vienna, whose works have been of immense service to the study of the Cynipidae, had previously suggested that the two species Pediaspis and Bathyaspis aceris were two forms of the same insect. His breeding experiments, like my own, have confirmed this hypothesis. diameter. But they are also found above ground where the root leaves the tree. When they form a cluster around a root, they give rise to a regular cylindrical swelling, which may measure 5 cm. or more in diameter. This gall consists at first of a firm somewhat fleshy tissue, with a small larva chamber, like that of Biorhiza aptera. When mature, it exhibits a wrinkled, woody, brown shell, of no great thickness, enclosing a large cavity. Experimental breeding. The galls mature in Spring, and are best collected in the month of March. They then contain the perfect flies, which have wintered in them. In the month of April they gnaw their way out and quit the gall. Fly. Length 5 mm.; colour, ferruginous; the face, the parapsidal furrows, a line at the base of the wings, the middle of the sternum, and the metathorax, blackish; the thorax slightly haired above, more strongly so at the sides; scutellum depressed, surface rugose, with a narrow smooth margin; abdomen smooth, slightly shining, the last segments darker above; legs reddish brown, strongly haired; antennae with fifteen joints. Experimental breeding. It is comparatively easy to rear this fly successfully. My own experiments were made last April. I began on April 12, having prepared for the purpose six small trees of Acer pseudo-platanus in pots. The flies placed on the trees began at once busily to examine the buds with their antennae, and prepared to prick them. Their method of setting to work was rather remarkable. When a fly had found a bud that suited her, she placed herself in such a position, head downwards, that she could drive her terebra diagonally from above towards the centre of the bud-axis. She pricked the bud several times, so that a number of eggs were deposited in the same bud. I could see with the aid of a lens the openings of the little canals pierced by the terebra on the covering of the bud; and an examination of the buds which had been pricked, showed that the eggs had been deposited on the rudimentary leaves. According to this, the galls ought |