762 PROFESSOR W. C. WILLIAMSON AND DR. D. H. SCOTT ON THE 
Longitudinal sections, however, show that the secondary phloém contained long 
parenchymatous cells in addition to the sieve-tubes. In the latter transverse walls 
are not found, and their ends appear to have been tapering (figs. 37 and 38). 
Generally, the larger elements appear to have been sieve-tubes, but the difference in 
size is not very marked. 
The preservation of the sieve-tubes is astonishingly perfect, so as to leave no doubt 
as to their nature, a point which was spoken of with some reserve in Memoir XIII. 
(p. 291), at a time when the detailed structure had not yet been observed. In 
several cases the sieve-plates on the radial walls of these elements can be recognized 
with certainty. This is the case, for example, at the place marked svup., in fig. 37, 
where the markings on the wall clearly indicate a compound sieve-plate. A much 
better example, however, is shown in the oblique section, from which fig. 38 was 
drawn. The sieve-tube in question is shown on a larger scale in fig. 884. The sieve- 
plates are marked on the wall in brown carbonaceous matter and exactly resemble 
those of some recent Gymnosperms and vascular Cryptogams. The smallest areas 
into which the plates are sub-divided no doubt represent the minute sieve-fields 
within which the actual pores were grouped. 
We may well wonder how structures so delicate, which are often difficult to 
demonstrate in recent plants, can have been preserved in the fossil state. That the 
markings in question really represent sieve-plates will not be doubted by any 
botanist who examines the sections for himself and sees how often these appearances 
occur, and how invariably they are limited to those elements which, from their position 
and form, must be regarded as sieve-tubes. Sieve-plates in fossil plants have 
occasionally been figured by previous writers ; among the most satisfactory are those 
shown by Messrs. BertRAND and RENAULT, in Poroxylon Edwarsu, which much 
resemble those of our plant.* 
It is not at all likely that minute differences in the thickness of a cellulose wall, or 
even minute perforations, would be recognizable so distinctly in the fossil condition. 
It is much more probable that what we see represents the carbonized remains of that 
portion of the contents of the tube, which adhered to the plate, so that the 
sculpturing of the latter has remained in the form of a carbonaceous print. 
Beyond the phloém there is a zone of tissue which we interpret as pericycle; it 
attained a great thickness, much exceeding that of the corresponding region either in 
Heterangium Grievit or in Lyginodendron.t It consisted, like the pericycle of those 
plants, of short-celled parenchyma, but contained in addition numerous short sclerotic 
cells either isolated or in groups (see Wuiti1amson, “ Organization,” Part XIIL, 
Plate 21, fig. 19), a feature in which this species differs from H. Grievii, where 
* © Recherches sur les Porozylons,” figs. 192 and 193. 
+ This layer was termed “inner cortex” in Memoir XIII., loc. cit., Plate 21, fig. 1; Plate 22, 
figs. 2 and 5, p. 
