56 BRITISH EOCENE FLORA. 



Non. S. Stcrnbergi, Heer, ' Flor. Tert. Helv.,' vol. i, p. 35, pi. xxi, fig. 5 ; vol. iii, pi. 2, figs. 1 — 4 ; 

 pi. vii, figs. 3 — 5 ; • Flor. foss. Arct.,' vol. i, pi. xxiv, figs. 7 — 9 ; vol. iii, pi. 2, figs. 1 — 4 ; 

 A. Sternbergi, Mass., 'Flor. Senigal.,' p. 154, pis. 5 — 7, 40; Sismonda, ' Mem. Reale 

 Accad. di Torino,' 180.5, p. 404, pi. iv, fig. C ! 



Upper part of Freshwater beds and Marine beds at Bourneinoutli. 



Only foliage has yet been met with in England. The leaflets are moderately short, 

 falcate, and awl-like, quadrangular in section, thickened at the base, and with the lower 

 side produced and decurrent on the stem, disposed spirally round the branchlets, which 

 they quit at first at a right angle and then curve gently upward and inward, free or in 

 slight contact at the apex with the row above. The branchlets, eighteen to twenty inches 

 in length, arc usually simple for five or six inches, and then fork copiously and chiefly 

 horizontally. They are articulated at the base, and when broken away a lozenge-shaped 

 scar remains, internally hard and ligneous, the woody core penetrating the smallest forks, 

 while the leaves spring from the external cortical system. The specimens either taper 

 and end in a leafy rosette or are constricted and swell into an oval bud. Branchlets 

 resembling these in the minutest particulars, articulated at the base, and representing one 

 year's growth, are annually shed by Araucaria Cunninghami, and are doubtless buried at 

 the present time on the muddy shores of the Brisbane River, very many miles away. 



In A. Cunniiirjliami the tough woody core of the branch penetrates half way into the 

 cone, when it gradually becomes spongy and vascular, and is lost towards the termination 

 in a pulpy mass. This is surrounded by a somewhat pulpy layer, a quarter of an inch 

 in thickness, a continuation of the fleshy bark from which the leaves spring. Upon this 

 the scales arc arranged spirally and at nearly right angles. When immature they are- 

 socketed into depressed scars on the axis, but, as the cone ripens, the attachment ceases 

 by desiccation, and the scales, to which the seeds are firmly soldered, become held only 

 by a single woody fibre which, parting from the central woody core of the axis, penetrates 

 the scale. When ripe and shaken by wind, the whole of the scales fall away at once, except 

 a few barren ones towards the base. The branchlets fall chiefly, I believe, in the spring, 

 long after the seeds have been shed. My friend, Leland C. Cossart, of INIadeira, has, 

 moreover, ascertained for me that foliage oli Araucaria Cunni'ngliami requires two or three 

 days to sink ; while of the mature seeds of a cone, none sank before five to six days, and 

 at the end of ten days there were still about 1 5 per cent, at the surface, so that in moving 

 water they would therefore necessarily be deposited separately. The axis of the cone 

 remains on the tree, for cones are never produced on the annual articulated branchlets, 

 but only on the persistent branches. Indeed, it appears that each compound scale of 

 the cone is an altered deciduous branchlet and not, as in the Abietinea?, merely an 

 altered leaf. Only in cases where some violence has operated could a young cone, with 

 the sap still flowing, and therefore in a condition to hold together, become detached. 

 These facts explain the absence of any trace of cones among the very large quantity of 

 ^ ' Verst.,' vol. ii, p. 204, pi. 39, fig. 4. 



