Fossil Cycadacem.—As in the case of vari ous unique types in the existing flora, the pres ent isolation of the cycads is connected with a past history of great extent. The living forms are only two lateral branches of a varied cycadalean group. The Cycadex are ancient, characteristic leaves of the Cycas type having already been present in the Carboniferous limestone. Undoubted petrified Cycas seeds occur in the Cretaceous of Vancouver Island. The 2amim were already present in the Jurassic. But they are always scantily present. Modern looking cycads appear to occur as certainly as anywhere in the upper Cretaceous (Atane Beds) of Greenland. Since Cretaceous time the northern limit of the Cycads has slowly retreated from the Arctic area to its present position. A species recalling the African Encephatartos was left behind in the lower Miocene of southern Europe. However, none of the fossil forms referable to the cycads have the importance attaching to the wholly extinct cycadeoid group, the Bennettitace.e. Nearly all the abundant vegetation of the Mesozoic, for merly supposed to fall within the Cycads, is being gradually relegated to this group, and to an account of its discovery and features we now turn.
II. Bennettitacete, The record of the dis covery and segregation of this extinct group from the cycads proper now forms one of the most interesting chapters of pala-ontologic re search. The cycad, or more exactly speaking, the cycadophytan type of foliage, characterizes the vegetation of mid-geologic time. Of all fossil plants yielded by the Mesozoic rocks, over 30 per cent are cycadalean. So universally and in such variety of species do the leaf imprints occur that the Mesozoic is often spoken of as the cage of cycads? But the actual proportion of cycad and cycad-like plants in the Mesozoic floras can only be surmised. The recovered record consists largely in xerophyllous, low land or swamp types, and it has not yet been ascertained whether there was a distinct cyca deous element in the contemporaneous upland and plateau vegetation; although it is probable that there was, and there is evidence that both cycads and cycadeoids were notable constitu ents in tfie Mesozoic vegetation of the Arctic areas. Probably the species ran into the thou sands. At any rate, the dominance of cycadeous plants in the fossil floras preceding the advent of recognizable angiosperms in the lower Cre taceous has long been regarded as a fact of deep evolutionary interest; though it is only within the past 25 years that much headway has been made in the study of the extensive fossil material at hand. Previously the knowl elge of the fossil cycads rested mainly on such facts as could be gleaned from the study of dissociated leaves and occasional stem imprints. The few petrified forms more accurately studied had not resulted in restorations, and it was naturally supposed that the extinct types were closely allied to modern forms. The names of the leaf genera, Zamites, Oiozamites, Cycadites, Stangerites, etc., etc., indicate this supposed affinity. But it is now known that even the xerophyllous types, which are.the more readily conserved because of form and habitat, often bore, despite their strictly cycad-like leaves, flowers and fruits very unlike the cones of existing cycads. (See Fig. 8). American re searches have mainly contributed to this result. Especially the discovery of the remarkable silicified trunks of the Black Hills of South Dakota has given to the cycadeoids a tran scendent interest from both the palwontologic and the botanic viewpoint. These trunks are. not, however, unique to North America. They belong mainly to the genus Cycadeoidea and the family Cycadeoidez, established nearly a cen tury since by the English geologist Buckland.
Various silicified stems with a heavy armor of leaf bases notably simulating the recent cycads just described had been encountered in the so-called "dirt bed" of the °Mite formation of the Isle of Portland. This lies just above the limestone there quarried. As the black earth of the "dirt bed" is stripped away, the cycadeoids, with more or less crushed terminal buds, are encountered quite in the original soil in which they grew. The stems are still being found and have been known to the quarrymen as "crowns nests" for several centuries, going back even to the time when the materials for the construc tion of Saint Paul's of London were quarried. Owing partly to the checked type of conserva tion, the Buckland specimens received little further study, and it was not until 50 years later that William Carruthers described an Isle of Wight trunk bearing ovulate cones.
Certain Italian specimens had yielded incomplete results, and a splendid Galician trunk in the Zwinger Museum at Dresden, discovered as early as 1753, remained unstudied. Neither did an interesting series of American trunks from the • iron ore" beds of Maryland, first noted about 1860, receive merited attention. In fact, the ovulate cone being the most aberrant cyca deoid feature, with the foliage uncertain, the relationship to the cycad-leafed plants of the Mesozoic was obscured rather than cleared; and so the subject rested until the attention of American palmontologists was directed to the Black Hills of South Dakota and a little later to the Freeze Out Hills of Wyoming. Begin ning with the species Cycadeoidea dacotensis (Macbride 1893) nearly a score of valid species have been described from silicified types, often bearing conserved crowns of young leaves and the flower buds in various stages of growth. In one instance, that of the monocarpic Cyca deoidea Dartoni, perhaps the most remarkable fossil plant ever recovered, fully 500 of the mature ovulate cones are embedded in the armor. Over half are complete and show the seeds with the finely conserved dicotyledonous embryos. For a full account consult references, Wieland,