CONTAINED IN NODULES OF LIMESTONE- 
153 
condition : the appearance is exactly that presented by the cellular tissue of Salishu- 
ria and Phyllocladus fruits, where the rich brown colour is probably due to a depo- 
sit of resin. 
Within these integuments is a large oval cavity, full of carbonate of lime and mag- 
nesia, of a yellowish white colour, and very compact. When the fossil is fractured, 
this mineral presents a cast of the surface of the cavity (as represented at Plate IV. 
fig. 1), and, when sliced, exposes a delicate membrane or sometimes two concentric 
ones, as represented in most of the specimens figured. These membranes have 
always more or less collapsed, and apparently are broken up into layers (Plate IV. 
figs. 7 & 8). These appearances of a double or treble membrane are probably due to 
the breaking up of one, or to the decomposition of the walls of the cavity ; for all are 
uniform in structure, and shreds of cellular tissue and scalariform vessels are often 
found uniting them, and they are further identical in structure with the walls of the 
cavity. A very highly magnified view of a portion, taken from the section figured 
fig. 7e, is given at fig, 15 , where the arrow indicates the position of the vascular 
bundles, which are more highly magnified at fig. 16 . At fig. 15 the section of this 
membrane is seen bent at an angle on the right-hand side, and traversing the trans- 
parent carbonate of lime ; to the left of it is a broad fissure in the mineral, and to 
the extreme left of the circle at f are the tissues of the inner wall of the second 
integument of the fruit. At fig. 17 is seen another section of the same membrane, 
formed of cylindrical utricles, with no traces of vascular tissue. 
We have been thus particular in describing these structures, because we find them 
to be uniform throughout the very numerous suites of sliced specimens which we have 
examined : it is scarcely necessary to add, that though they exhaust our materials, they 
leave much to be desired in our knowledge of the fossil to which they belong. Nothing, 
however, has occurred during our study of them, to warrant our expecting to find any 
further structure in the cavity of the seeds taken from the limestone nodules. 
Although no positive proof of the real nature and affinities of Trigonocarpon can be 
offered until the discovery of embryos or spores within the cavity of the fruit, there are 
so many important points now shown to exist in their structure, as to warrant more 
exact comparisons than have ever yet been instituted. We have already mentioned 
Saiishuria as the nearest existing analogue known to us, and shall accordingly pro- 
ceed to discuss it. 
Saiishuria is a drupe-bearing coniferous tree, a native of China and Japan, long 
cultivated in Europe, but only producing fruit in the middle and southern regions of 
this continent; these are oblong, about the size and colour of a damson plum, and 
are produced on the terminal branches of the tree, from which they are easily de- 
tached when ripe. The drupe consists of three integuments, which are the meta- 
morphosed coverings of a naked ovule. The external integument is thick and fleshy 
as in Trigonocarpon, is covered with a delicate cuticle (Plate V. fig. 6), and is 
formed of membranous utricles; the outer layers of these cells are empty, but the 
