GYMNOSPERMiE. 125 



tines, but the ovule in Taxus subsequently develops solitarily on its cup-like base, whilst 

 the support in the Abietincae takes a scale-like form. A primordial member of the Abie- 

 tinea?, before the ovules were collected in a cone, should not differ fundamentally in its 

 fruity organs from a primitive Taxad, before its floral axis had been impoverished. While, 

 as already stated, it appears certain that the Taxese as a group preceded the true Coniferae, 

 it is probable that what are now alike cone-bearing Coniferae were differentiated before 

 the fruiting organs of some of them definitely took the form now common to all. The 

 cone is a branch wholly or partially modified to serve as a support to the female organs, 

 and to protect the seeds after fertilisation until they ripen. In the cones of Cycads the 

 metamorphosed leaves directly support the ovules ; but in Coniferae these organs are 

 situated on structures arising out of the axils of the bracts, which latter do not increase 

 while the ovule-bearing structures are enlarging and producing the ovules. In the Abie- 

 tineae an entire branch is modified to form the cone, but in the Cupressineae only the 

 upper part is transformed. The occasional perfoliate cones of Cryptomeria and Cunning- 

 hamia demonstrate that the middle of the branch was the part originally modified, and 

 that the sterile termination has since been lost through atrophy. In many of the older 

 Coniferae of the Trias and Rhaetic the scales of the cones are more scattered and their 

 leaves less completely altered, producing the long cylindrical forms which so often charac- 

 terise primitive types such as Voltzia, Schizolepis, Glyptolepis, and Lepidostrobus. The 

 most ancient Conifers of which we have definite knowledge, although there must have 

 been still older types, are the Walchia, a considerable tribe of trees of large growth, 

 clothed with deciduous foliage not differing greatly in appearance from that of Araucaria 

 Cookii and A. excelsa. The cones, however, were small, ovate or oblong, with lanceolate 

 scales formed of only slightly modified and closely imbricated bracteal leaves firmly 

 attached to the axis. The seeds were small, slightly alate, and free, from one to three being 

 born under each scale. In construction the cone is not unlike that of Araucaria, except 

 that the latter possesses but one seed to each scale, and the bases are not persistent to the 

 axis. The Walchice are associated in the Carboniferous with two genera of Ginkgos, Ginkgo- 

 phyllum and Trichopitys, together with the last of the Cordaites. They are succeeded in 

 the later Permian by TJllmannia which links them to the Jurassic Br achy phyllum. Voltzia 

 of the later Permian and Trias is seen, by the disposition of its seeds, and in other respects, 

 to be allied to the Taxodieae, though its thin imbricated scales, forming a loose and rather 

 large cone, link it also with the Araucarieae. The Triassic Albertia bears a striking resem- 

 blance to Agaihis in its cones, but in foliage it more resembles Araucaria Bidwillii and 

 A. Cunninghami. Glyptolepis, a genus bearing cones with a long axis and small fim- 

 briated scales, carries on the Voltzian type through the Upper Trias. 



The commencement of the Jurassic period marks an epoch in the development of the 

 Coniferae, their characters for the first time approaching sufficiently to those now existing 

 to enable the whole to be placed in living families. At the same time many of the archaic 

 genera, such as Walchia and the Sequoia-like Palissya, do not survive beyond it, but make 



