232 Browne.—A Second Contribution to our Knowledge of the 
transitional from a large cone, Cone D, to the fertile stem was also cut 
serially, and proved abnormal in several respects. Serial longitudinal 
sections were made of a normal cone, E. Besides some hand-sections, 
serial sections of portions of four normal cones were cut, but as their study 
only confirmed the conclusions from Cones A and B no description of them 
seems necessary. Serial sections of a mature node of a sterile stem were 
also prepared for comparison with the fertile stem below Cone D. 
Anatomy of the Cone. 
The longitudinal reconstructions of the xylem in Cones A and 
B (cf. PL XII and PL XIII) show that the whorls of traces are not 
inserted truly horizontally on the axial stele ; this appearance is not due 
to the obliquity of the sections, as the appearance of the individual 
tracheides is that of the structures cut horizontally. The insertion, both 
of the traces and of the sporangiophores, is often distinctly irregular, but 
much more so in the large Cone B than in Cone A. In the lower region, 
where the elongation of the internodes is naturally greater than in the upper 
part, it is often easy to recognize the traces that belong to any one whorl. 
In Cone A there seem to have been nineteen whorls, excluding the terminal 
cluster of incompletely differentiated sporangiophores. In Cone B, owing 
to the much greater irregularity in the disposition of the traces and of the 
sporangiophores, it is very difficult to estimate the number of whorls 
present. The impossibility of saying with any certainty to what whorl 
many of the traces belong makes it difficult to say of what order are the 
parenchymatous meshes arising above some of them, or, in other words, to 
say into what number of internodes these meshes extend. Thus, in Cone B, 
many of the whorls of traces are duplicated over a portion of their extent; 
when this duplication takes place freely, we have what may be termed 
a pseudo-whorl; that is to say, while the increase in the number of the 
traces in such a region, compared with the number of traces in other 
whorls in that part of the cone and with the diameter of the stele, is too 
great for us to regard this phenomenon as merely an unimportant variation 
leading to a slight increase in the number of traces in a whorl, the number 
of supernumerary traces that have been called into existence is not suffi¬ 
ciently large for them to constitute an independent whorl strictly equivalent 
to other whorls in that part of the cone. I estimate that there are about 
twenty-eight whorls in Cone B, three of which are pseudo-whorls. Were 
lines delimiting these whorls included in PL XIII they would be very 
sinuous and irregular; as, however, their course would largely depend on 
individual interpretation, they have not been marked. A less great 
increase in the number of traces of a whorl, though an increase leading to 
the development of traces, situated, as are those of a pseudo-whorl, a little 
above other traces of the whorl, I have considered merely as a local dupli- 
