142 BOTANICAL GAZETTE [aveust 
It is noteworthy that in certain of the fossil Lycopodiales, as above 
“mentioned, there is clear evidence of the possession of mesarch struc- 
ture in the primary wood of the leaf-trace. 3 
That there is no sharp line of cleavage between mesarch and exarch 4 
vascular cryptogams is thus perfectly clear. Among the main groups, — 
we have numerous instances of both types of development, and the 4 
one merges, little by little, into the other. In Lycopodium, the — 
centrifugal wood of the stem, when present, consists of only a few — 
tracheids; in such forms as Lyginodendron, for instance, and in — 
many ferns, the spiral elements are near the outside, while in 4 
Tmesipteris and Phylloglossum, they are in the very center. The 
variability of the position of the protoxylem in the primary wood 
bundle, and its consequent unreliability as a phylogenetic character, — 
cannot be too strongly emphasized. . 
While the structure of the primary bundles of the vascular cryP 
togams cannot be used as a guide to their interrelationships, t 
furnishes an excellent character for the whole group, both living and a 
fossil forms. ‘The presence in the stem of centripetal primary wood, 4 
continuous with the protoxylem, is the distinguishing mark of these — 
plants. The older French botanists appear to have been perfectly 4 
right in characterizing their bois centripete as cryptogamic wood, bie! E 
the Sphenophyllales, Lycopodiales, and Filicales show this structure a 
very clearly. The only apparent exception? to the rule is the cala- a 
mitean series, including the modern Equisetales. A species of Cala- 4 
mites, however, the so-called C. pettycurensis, or Protocalamites, has “ 
recently been shown to have well-developed centripetal wood on the a 
inner face of the protoxylem canals of the stem (Scott 7). Further, a 
EaMEs (10), working in this laboratory, has distinguished a mesarch 3 
condition, with consequent presence of centripetal xylem, in the etl” 4g 
of the vegetative and reproductive leaves of several living specie . 3 
Equisetum. It thus seems certain that centripetal wood was onC? — 
well developed in the Calamites and their allies, but in the proces 
of time has been almost entirely lost. t 
It is worthy of note that though secondary wood is often pres 
in the vascular cryptogams, and may constitute the bulk of the cen" 
cylinder, the protoxylem is intimately associated with the centtip” 
2 Exclusive of that aberrant group of ferns, the Ophioglossales. | 
