A FOSSIL FLOWER FROM THE EOCENE. By Edward W. Berry, Of the Johns Hopkins University, Baltimore, Maryland. Flowers are by no means so common in the fossil state that new occurrences are without exceptional interest. If the flowers preserved in the Baltic and other amber deposits are omitted from the enumera- tion of this class of remains their number is very limited and their preservation is often not all that could be desired. For these reasons I am prompted to publish a preliminary description of a weU-preserved flower from the Eocene of southern Tennessee collected some years ago by Prof. L. C. Glenn, of Vanderbilt University, and preserved in the collections of the United States National Museum. I have been engaged for some years past under the auspices of the United States Geological Survey in a study of the wonderfully rich fossil floras of the Southern States, particularly that of the Lower Eocene, or Wilcox group of formations. The latter flora as known at the present time includes upwards of 250 species of subtropical, largely strand, types. Most of the species are new and the flora as a whole is unlike described American Eocene floras which come largely from the Rocky Mountain province. It is comparable with those of the European Eocene and with the existing flora of the West Indies, Central and northern South America. The specimen described in the following note is made the type of a new genus, since it is not referable with certainty to any of the existing genera of the family with which it shows the most affinity. COMBRETANTHITES, new genus. This genus is proposed for the following species which is based on a fossil flower referable to the Combretacese and very similar to the flowers of some of the species of the genus Corribretum. To avoid any seeming inaccuracies the species is described in detail from the speci- men even though this repeats some of the floral characters that run through the family. Proceedings U. S. National Museum, Vol. 45— No. 1980. 261 262 PROCEEDINGS OF THE NATIONAL MUSEUM. VOL. 45. COMBRETANTHITES EOCENICA, new species. Peduncle stout, curved, about 4 mm. long. Calyx rather deeply 4 or 5 lobed, the lobes ovate in outline and with bluntly pointed tips. Corolla polypetalous of 4 or 5 petals alternating with the calyx lobes, long and narrow, seemingly pointed, about twice the length of the calyx lobes. Ovary inferior, style long and slender, probably with a single terminal stigma. Stamens 12 in number with long slender filaments, exserted. Anthers elongate elliptical, two-celled, dehiscing by longitudinal slits. The stamens may vary in length or their apparent variation may be simply a feature of preservation. The present species is based on the exceptionally preserved flower shown natural size in figure 1 of the plate, lying across a leaf of Cassia emarginata Berry, the other markings on the leaf being those of a well-marked leaf-spot fungus. It is also to be pointed out that figures 1-3 are from photographs that have not been retouched in any particular. I am not sure that appearances that I have interpreted as petals are correctly identified, but it is hard to imagine what else they can possibly represent. The single slender style is also a feature that may be simulated by a filament. As shown in the accompanjdng res- toration the flower is polypetalous regular and perfect. It is represented as having a four-lobed calyx and four petals, although only three calyx lobes and two petals are dis- tinctly seen in the specimen. If four is the cor- rect number, then the stamens are three times as numerous as the petals. The reason for considering that these flowers were capitate or in crowded spikes is their small size, narrow petals, and exserted style and stamens — all characters shared by the flowers of the Mimosacese and Combretacese, the two families whose flowers are most like the fossil. In the Mimosacese the filaments are usually more slender and more elongated as well as more or less united, while they are free in the fossil. The anthers are also much smaller and less elongated in the Mimosaceas. Most of the Combretacese have flowers very similar to the fossil although the stamens are usuaUy reduced in number to twice the number of the petals or of the calyx lobes in the apetalous forms. However, some of the modern forms have thrice as many stamens as petals or calyx lobes. The most similar modern flowers in appearance that I have been able to find are those of Comhretum guanaiense Rusby, from Bolivia, and in this the stamens are only eight in number and more exserted. I Fig. 1.— Restoration of com beetanthites eocenica Enlaeqed four times. NO. 1080. A FOSSIL FLOWER FROM THE EOCENE— BERRY. 263 have submitted the specimen to various botanists famihar with the flora of tropical America and compared it with a vast amount of recent material and am satisfied that it represents an Eocene member of the Combretacese, a family that was apparently well represented in the early Eocene, since I have described (in manuscript) from contempo- raneous deposits the leaves of two species of Combretum, two species of Terminalia, one species of Conocarpus, and both the leaves and fruit of a species of Laguncularia. These all serve in a measure to substantiate one another and a certain amount of confirmatory evi- dence is furnished by the petrified wood described by Felix from the European Eocene as Combretacinium ^ and compared with the woods of modern forms of Terminalia, Bucida, etc. Leaves of Terminalia and Combretum have also been described by various authors from the European Tertiary, tending to show the great similarity between the flora of Europe and that of southeastern North America in Eocene times and the tropical American character of European Eocene floras. Formation and locality .—Wilcox Group of the Eocene, 1^ miles west of Grand Junction m Fayette County, Tennessee. (Collected by L. C. Glenn.) Holotype.— Cut. No. 34445, U.S.N.M. EXPLANATION OF PLATE 21. Fig. 1. Impression of Combretanthites eocenica, new genus and species, lying across a leaf of Cassia, natural size. 2. Same enlarged four times. 3. Same enlarged six times. 4. Same as figure 3, photograph retouched. Figs. 1-3 are from photographs without retouching in any particular. I Felix, Zeits. deutsch. geol. Gesell., vol. 46, 1894, p. 90, pi. 10, figs. la-c.