io6 Lady Isabel Browne . 
only appear to be centripetal, but that they have no protoxylems 
and therefore show no signs of fascicular individuality. The third 
suggestion, namely that these lateral strands represent a fresh 
development, is untenable in view of the conclusive evidence that 
the Equisetaceas have in the course of their evolution undergone 
considerable and continuous reduction. It therefore seems that 
they must either represent reduced secondary xylem, as suggested 
by Dr. Scott, or reduced centripetal primary xylem as contended by 
Mr. Gwynne-Vaughan and Professor Bower (4) (2). Against the 
former alternative it may be urged that the development of these 
lateral strands appears to he centripetal; that Mr. Cormack has 
discovered what he considers to be secondary thickening at the node 
of Equisetum maximum (3) and that this thickening appears to be 
unconnected with the lateral strands; but this so-called secondary 
thickening is compared to the nodal thickening of a Calamite, and 
as Dr. Scott pointed out, in the course of lectures previously 
referred to, it has not been proved that this nodal thickening of 
Calamites was a secondary growth. If, however, Dr. Scott is right 
in regarding these lateral strands as reduced secondary xylem their 
lateral position is unintelligible, as secondary xylem naturally arises 
at the periphery of the primary xylem. Against the view that the 
lateral strands represent primary centripetal xylem it may be 
pointed out that these lateral strands do not occur in any of the 
numerous and well-preserved Calamarise; as these plants are 
probably very close to if not actually on the direct line of descent 
of the Equisetaceae, we should expect that, if the lateral strands 
represent vestigial centripetal xylem, they would be at least equally 
well represented in them. Further, centripetal xylem is only known 
among the Calamariae in Calamites petty curensis, where it occurs on 
the inner edge and not at the sides of the carinal canal. But the 
whole of the axis within the carinal canal was presumably at one 
time in the history of the race occupied by centripetal xylem, i.e., 
the Equisetales probably had protostelic ancestors. The reduction 
of the centripetal xylem, reduced to an arc in C. petty curensis , was 
probably brought about by the development of centrifugal xylem, 
which, as observed by Dr. Scott, frequently encroaches upon centri¬ 
petal xylem (14). As the development of the median strands 
commences from a median point, and as they are wedge-shaped, it 
may well be that the arc of centripetal xylem, present in C. petty- 
curensis, was split by encroachment of the centrifugal xylem and 
that this process resulted in the lateral, apparently centripetal 
strands of the Equisetaceous bundle. On the whole, however, the 
