It has recently been shown that the Permian or Permo-Car boniferous vegetation of some regions in China was of the typical northern type. Prof. Halle of Stockholm in an exceptionally im portant memoir (1927) has given an account of the late Palaeo zoic floras of central Shansi which are preserved in two series of sediments; the older strata known as the Yuehmenkou series and above it the Shihhotse series. Both floras included species of Calamites, Sphenophyllum, Pecopteris, Callipteridium, Cordaites, Stigmaria and other northern Permo-Carboniferous types. As in the Kusnezk flora so also in the collections from the Shihhotse series there are species belonging to genera such as Cladophlebis, Neuropteridium, Chiropteris and Dioonites which bear a very close resemblance to members of Triassic floras in many parts of the world. Chiropteris is a genus represented by broad leaves with forking and anastomosing veins which in form remind one of the fronds of some species of the fern Ophioglossum; its systematic position is uncertain. Specimens are recorded from Triassic beds in Germany, South Africa and elsewhere. The name Cladophlebis is given to the branched fern fronds bearing comparatively short and often slightly curved leaflets, with a central rib and arching, forked lateral veins, attached by the whole of the base ; species are characteristic of many Mesozoic floras but are seldom met with in Palaeozoic floras. The Shansi floras differ from the Kusnezk flora in the absence of Glossopteris and other Gondwana land plants. The Shihhotse series is overlain by barren sedimen tary beds indicative of desert conditions ; it is highly probable, as Halle suggests, that in these beds we have evidence of a change from a genial and humid to a dry climate, which coincided in date with a corresponding revolution chronicled in the Permian period over a wide area in the northern hemisphere. Floras similar to those of Shansi are recorded also from Korea.
This incomplete presentation of facts may serve to illustrate some of the broad features in the distribution and composition of floras at the stage of geological history immediately antecedent to the Mesozoic era. There were at least three botanical provinces characterized by floras that were distinguished by the abun dance of certain genera: (I) A very widely distributed and fairly homogeneous northern flora which probably had its origin in the lands north of the Tethys sea and spread southward into China and members of it reached the Malayan region as invaders of the realm of Gondwanaland; (2) the Glossopteris flora from which in late Permian days some of the plants wandered far to the north, and Glossopteris penetrated to Greenland well within the Arctic circle; (3) the Kusnezk flora of Upper Permian age, which was in part composed of plants which had found their way across the Tethys sea from the south and in part of representatives of north ern Permian floras.
At the beginning of the Mesozoic era the prevalence of desert conditions over a wide area in North America and in the Old World was highly unfavourable to the further development or the continued existence of many members of the rich Permo-Carbonif erous flora. By far the greater number of the Palaeozoic species
failed to survive; a few new forms were evolved, and some plants such as Schizoneura and V oltzia were immigrants from Gondwana land. A comparison of late Palaeozoic and early Mesozoic floras reveals a sharp contrast which furnishes an interesting illustration of the interdependence of organic and inorganic evolution, of the effect of changing geographical conditions on the evolution of the plant world. One of the well established conclusions of geologists is that there have been recurrent cycles of mountain-building and, as a necessary consequence, recurrent interferences with the fac tors conditioning plant life. One of these geological revolutions oc curred at the end of the Palaeozoic era not only in the northern hemisphere but also in China and in some other regions. One of the more important aims of the student of evolution is to discover connecting links in the plant world, particularly at epochs of widespread crustal disturbances. Attention has already been called to the presence in late Permian floras of genera that are character istic of Mesozoic floras : types, which became prominent in the lat ter part of the Triassic period and persisted through the several stages of the Jurassic period, were already in existence, though in a subordinate position, before the close of the Permian period. The greatest contrast between the two eras is the marked differ ence in the general facies of the vegetation consequent on the disappearance of most of the commoner members of the older forests ; but there is also a considerable difference which is due to the relative abundance in the Mesozoic floras of plants of more modern aspect, plants that are closely akin to living species as con trasted with the more archaic and much less familiar types in the Palaeozoic floras. A comparison of the rich fern vegetation of the Upper Triassic and Rhaetic floras with the comparatively small number of true ferns in the Permo-Carboniferous vegetation raises a question which has not been fully answered ; whence came the more modern types of ferns which characterize the Mesozoic floras? The late Palaeozoic ferns, apart from the Marattiaceae, which can be linked with Triassic and Rhaetic forms are very few, and the evidence they afford of direct relationship to Mesozoic ferns is not convincing.