Summary. The chief points to be gathered from the investigation into the subject of this paper are the following:— 1The universal presence of transfusion-tissue in the leaves of Gymnosperms whose venation is of a reduced or rudimentary character, and where the complex reticulate or dichotomous venation of the Ferns or the higher Angiosperms is lacking. 2The conspicuous development of centripetal xylem in the vascular bundles of the foliage-leaves of all modern Cycads, without exception, where it constitutes the most important part of the xylem of the bundle. This is a point which has long been familiar to botanists. 3The discovery, for the first time, of a mesarch structure of the vascular bundle in the leaves of Coniferae. In the cotyledonary bundles of Ginkgo biloaa, Linn., where this was first observed, the centripetal xylem forms the chief part of the woody strands and is very conspicuously developed. In the bundles of the cotyledons and mature leaves of all other Coniferae which were examined the centripetal xylem is extremely reduced, this reduction having, in most cases, corresponded with a greater development of the centrifugal xylem. It is most obviously present in those genera which approach nearest to Ginkgo in general relationship, such as Cephalotaxus and Taxus. In many genera, however, no trace of this tissue could by discovered, at least in the sections examined, though it is not improbable that isolated remnants of it may occur here and there along the course of the bundle through the leaf. 4The occurrence, in the cotyledonary bundles of Ginkgo biloba, Linn., and Cycas revoluta, Thunb., of an intimate union and gradual transition between the tracheides of the centripetal xylem on the one hand and those of the transfusion-tissue on the other; whence it may be justly inferred that the latter tissue owes its origin, wherever it may subsequently be found, directly to the former. Though in the leaves of the majority of Coniferae this origin is by this time almost entirely obscured, owing to the more or less complete obliteration of the centripetal xylem, it is still often evidenced by the presence of transfusion-tissue on the ventral side of the xylem, by the very frequent extension of the lateral transfusion-tissue towards the ventral side of the bundle, and by the transitions in the characters of the tracheides between those most externally placed and those nearest the protoxylem. In the foliage-leaves of Cycas the connection bet?veen the two tissues is obvious. 5The final inference, that transfusion-tissue, which occurs almost universally in the leaves of Gymnospermous plants as an auxiliary conducting-system, has been phylogenetically derived from the centripetally-formed xylem of the vascular bundle, and is thus, morphologically, an integral portion of the bundle itself. I am indebted to Mr. C. P. Robertson-Glasgow for the use of some excellent preparations of the cotyledons, leaves, &c., of various Coniferous plants. My best thanks are also due to Dr. D. H. Scott for many invaluable criticisms and much kind assistance during the progress of my work. [The important discovery, quite recently made by the two Japanese botanists, Hirase and Ikeno, of spermatozoids in the pollen-tubes of Cycas revoluta, Thunb., and Ginkgo biloba, Linn., and still more recently by Webber in a species of Zamia, bridges over in a striking manner the gulf, hitherto supposed to exist between flowerless and flowering plants, between Ferns or Fern-like plants and Gymnosperms. The relation between the two great groups of plants had already been partly indicated by the Fern-like foliage of such a Cycad as Stangeria. One of the most important of the facts brought forward in the present paper, viz. the occurrence of centripetal xylem in the cotyledonary bundles of both Ginkgo biloba, Linn., and Cycas revoluta, Thunb., seems to me (especially in view both of the external characters and anatomical structure of such fossil forms as the Medulloseae, Lyginodendron, and Nggerathia) to strongly support the above conclusion that Ginkgo and the Cycads hold an intermediate position between some primitive group or groups of Fern-like plants and modern Gymnosperms.