42 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



additional covering round the ovule, which French botanists, especi- 

 ally, regard as a rudimentary carpel. But, as yet, no transitions are 

 known between this elementary open condition and the closed carpel 

 of all existing Angiosperms. All Gymnosperms leave the summit of 

 the nucleus of the ovule exposed, so the pollen falls directly upon it. 



Cordaiteae is the name given to a family of fossil plants which bore 

 stamens and ovules. " They were tall, somewhat slender trees, with 

 trunks rising to a great height before branching, and bearing at the 

 top a dense crown composed of branches of various orders, on which 

 simple leaves of large size were produced in great abundance." * 



The veins of the leaves are repeatedly forked, unlike those of our 

 timber trees (excepting the Pines), in which they are " reticulated " or 

 " netted." 



Cordaianthus is a genus of which the flowers have been preserved. 

 They are in catkins or cone-like structures, consisting of bracts or scales. 

 The males had one or more stamens in front of each. The females 

 formed similar catkins, each scale having an ovule in front of it. 



Although no true links are known between the cup-shaped, so-called 

 " carpels," yet we can find some resemblances, if nothing more, between 

 certain angiosperms and the fossil seed-plants. Turning to the catkin- 

 bearing families, these are probably relics of much older and larger 

 families, as indicated by their having so few genera at the present 

 day. They are the Cupuliferae (lo), the Salicaceae (2), Plataneae (i), 

 Juglandeae (5), Myricaceae (i), 2in6.Casuannaceae (i). Cupuliferae (10) 

 has three tribes : the first has two genera (Birch and Alder) ; the second 

 has four genera, including the Hazel and Hornbeam ; the third has 

 four genera, including the Oak, Beech, and the edible Chestnut. Lastly, 

 the Salicaceae (2) contains only the Willow and Poplar. 



The flowers are quite as simple as those of the fossil genera ; the 

 males consisting of a scale or bract with one or more stamens, the 

 females of one or more carpels. 



It is noticeable in the figure of Cordaianthus, as given by Dr. D. H. 

 Scott (a genus of the group Cordaitales intermediate between Pterido- 

 sperms and the mesozoic Cycadophyta, leading up to our existing 

 Cycads), that the anther-cells are distinct on a common filament. This 

 perhaps may be attributed to the " forking " method of branching of 

 the vascular bundles, so common in fossil plants, as in the leaves, still 

 to be seen in the Ginkgo-tree {Salishuria adiantifolia ; so called because 

 the leaf strongly resembles the leaflets of the Adiantum or Maiden- 

 hair Fern). 



A similar separation in different degrees of the pollen-sacs or anther- 

 cells is seen in several of the Cupuliferae, as the Hazel and Hornbeam, 

 Myrica and Casuarina. The anther-cells are also more or less distinct 

 in the gymnospermous Coniferae, as of Pinus and Araucaria, &c. 



With regard to the insertion or position of the ovules upon the 

 carpel, it is familiar to all that they grow from the united margins of 

 the carpellary leaf, as seen in a pea-pod. The vascular cords running 



* Scott, op. cit, ii. p. 519. 



