750 TRANSACTIONS OF SECTION K. 



in the adjacent hypoderm. Probably they (1) are organs of respiration, and 

 (2) render the absorption of water more efficient and controllable in a water- 

 logged soil, such as I), alliaceum Bl. affects. The walls of the endoderm are not 

 specially thickened ; sooner or later they contain a suberised lamella. 



Secretion cells occur in the cortex in all except the Cedreln spp. ; but even in 

 these their absence cannot be positively affirmed, as the material had lain in 

 alcohol. 



An endotrophic mycorhiza, which appears to degenerate and to be digested in 

 the deeper cortical cells, is present in more than half the species ; in addition an 

 exotrophic mycorhiza occurs on some of the rootlets oi Sandoricum luciduni. 



In plants like the beech, which have thin much-branched rootlets, the total 

 length of the rootlets and the total surface they utilise for absorption in a given 

 volume of soil are greater than in plants like the ash, whose rootlets are thicker 

 and less branched. The thin rootlets of the beech are in a better position to 

 penetrate among the finer particles of the soil, and thus exhaust their water of 

 hygroscopicity to the last drop. The beech is fitted for an intensive, the ash for 

 an extensive, system of work. 



The root system of Meliaceae is fitted for extensive rather tban intensive work. 

 The rootlets are, relatively speaking, very thick and very scattered. The hairless 

 rootlets of the Dysoxylon spp. in particular are strikingly thick, and recall those of 

 a saprophytic plant. 



The explanation of all appearances in the anatomy of the epiderm, hypoderm, 

 and endoderm, and of the character of the root-system as a whole, whether inten- 

 sive or extensive, must be souglit for in the texture, composition, and continual 

 or periodic dampness of the soil, in the leaf-arrangements for transpiration, and in 

 the predisposition of the plant itself. 



The work was carried out under the direction of Professor Biisgen in the 

 Forstakademie, M linden, Hanover. 



2. The Structure and Wound-reactions of the Mesozoic Genus Brachy- 

 phyllum. By Professor E. C. Jeffrey. 



The affinities of Brachyphyllum have been much disputed since Brongniart first 

 described it as a 'coniffere douteuse.' From material with structure preserved 

 by partial charring, from the Earitan deposits of Staten Island, N.Y., U.S.A., and 

 representing the Brachyphyllum macrocarpwn of Newberry, it has been possible to 

 settle that the genus is an Araucarian Conifer of Cupressinoid habit. It differs 

 superficially from any living representatives of the Araucarinese by \ta flattened 

 branches and adnate leaves. The central cylinder encloses a much smaller pith than 

 do those of Araucaria and Agathis, and the pith is occupied by masses of scleren- 

 chyma. The wood is of a more primitive type than that found in the still existing 

 Araucarineae, resembling the ancient Gymnosperms in the absence of wood 

 parenchyma. The pits of the tracheids and of the rays are of the Araucarian type. 

 The phloem is without bast-fibres, and resembles in this respect the Abietinese. 

 The leaf-traces pass off to the leaves with a well-marked foliar gap, and fork 

 repeatedly in the leaves, as in Agatliis, both features of diflerence from any known 

 representatives of the Lyeopods. No indications of the presence of ligular struc- 

 tures such as are found in some of the Lyeopodinese have been made out. The 

 ramifications of the leaf traces become surrounded by transfusion tissue, resembling 

 that found in many of the Abietinese ; for the phloem is completely included by 

 the tracheidal cells. The transfusional borders of the bundles fuse in the upper 

 parts of the leaves, forming a continuous zone above the active parenchyma, which 

 lies on the morphologically lower (outer) surface of the leaf, since the upper 

 surface is adnate to the branch. 



Araucarian cone-scales found in constant and intimate association with the 

 branches described above and closely resembling those from Cretaceous deposits of 

 Greenland, Europe, and the United States, referred by Heer and others to the 

 living genus Dammara, are shown not to belong to this genus by the fact that they 



