Admirable as this scheme is, and scientifically acceptable as it has proved to be, it nevertheless presents certain well-recognized defects with respect to the requirements of theory, although at the time of its formulation and as late as 1899 it represented the sum of available knowledge. It was just at this time that paleobotany became available as a means of meeting those deficiencies which a knowledge of living plants could not overcome. For a long time botanists have been familiar with certain Paleozoic remains having a fern-like aspect which were generally accepted as ferns; but because of their want of direct connection with stems or fruit, there remained a serious doubt as to their real character. In the same horizons, detached fragments of stems were also observed with increasing frequency. The study of their anatomy disclosed a structure which, in some respects, was curiously like that of ferns, while in other respects it approximated to the anatomy of the higher plants as presented in some of the gymnosperms. This combination of filicinean and cycadean characters was noted by Potonié, who succeeded in correlating them and expressing their phylogenetic position in the name of a new order which he called the Cycadofilices.
There yet remained to be considered certain remarkable fruits for which no relationship has as yet been determined until, through the work of Scott, Oliver, Kidston and others, it was shown that they were of the nature of seed-bearing organs which could be correlated with the Cycadofilices. It thus became evident that there was a hitherto unknown group of plants combining the characters of ferns in their foliage and stem structure with those of primitive gymnosperms as presented in their stems and fruits. On the whole, however, these plants approached most nearly to the pteridophytes in their external features. To this new phylum, of which the Cycadofilices formed the most conspicuous member, Scott and Oliver in 1904 assigned the most appropriate name, Pteridospermæ. This result was based entirely upon paleontological evidence through comparative anatomy, and it compels us to recognize the existence of five, instead of four great phyla. The farreaching significance of this achievement can not be overestimated. It is not only of the utmost importance as proving the general course of evolution and bringing into the realm of proved facts what had previously been a working hypothesis only, but it offers an entirely new point of departure for the botanist of the future. Attention may also be directed to one other effect. The tendency of this discovery is to coordinate, unify and strengthen all branches of botanical knowledge, bringing to us the conviction that the more extended and thorough our knowledge of the earlier forms of vegetation becomes, the more satisfactory will be our knowledge of the science as a whole; for while the example selected is probably the most important for our special purposes, the general utility of paleontological research in relation to the
