598 
MR, F. O. BOWER ON THE COMPARATIVE MORPHOLOGY OF THE 
the cells of the internal tissue at the extreme apex of the leaf contain single, large, 
compound crystals apparently of calcium oxalate ; these remain permanently as the 
leaves develop, and point to an early cessation of meristematic activity in the tissues 
at the apex of the leaf. The tissues below this apex of the conical phyllopodium 
retain their activity and growth, which go on almost uniformly for a time over the 
whole of any given tranverse section. Soon, however, differences appear in this 
X 
respect; the cells of the dorsal side of the phyllopodium cease to divide in longi¬ 
tudinal planes, though longitudinal sections show that most of the elements continue 
to divide by tranverse walls : at the same time the dorsal portion increases in bulk, 
while the tissues become differentiated, and show all the signs of passing over to 
permanent tissue. Along the ventral surface of the leaf, however, and especially at 
the two corners of its outline of transverse section, which is now angular, the cells 
still divide in longitudinal planes, and show signs of great meristematic activity : the 
division-walls, which appear at the corners of the section, show considerable regularity 
of arrangement. All those divisions of this period which can be seen in a transverse 
section are anticlinal: the result of this, together with the accompanying growth, is 
the formation of two projecting wings attached at the corners of the ventral side of 
the phyllopodium (Plate 40, figs. 49, 50), and extending from close below the apex, 
which is itself not winged, almost to the base. Since all the divisions of cells in the 
wings are at first anticlinal, the tissues develop as well-defined layers, parallel to the 
outer surfaces ; each wing consists of upper and lower layers of dermatogen, which, 
as usual, cover the surface externally, and enclose eight or nine regularly arranged 
layers of internal tissue. The growth of the wings is not localised at any definite 
point, but the divisions at the margin are less frequent than nearer the midrib : this 
shows that the growth is intercalary, not marginal, as in the Ferns. The uniformity 
of arrangement of the divisions in the wings is disturbed at certain points by the 
appearance of periclinal and irregular divisions in the central layers, which result in 
the formation of procambial strands. The periclinal divisions also extend in some 
cases, where a large bundle is formed, to the layers immediately below the derma¬ 
togen ; the cells thus produced, together with the vascular bundle which they enclose, 
form one of those ribs which project on the lower surface of the mature wing. 
Applying the same terminology as has been used above, it is obvious that the 
leaf of Gnetum Gnemon is a simple unbranched phyllopodium, on which wings are 
developed in a corresponding position to those in the Ferns and Cycads : there is no 
definite peculiarity of conformation at the base of the phyllopodium, beyond a slight 
increase in bulk towards the base, which assumes the form of a smooth transverse 
ridge on the ventral face, above the position of those glandular structures which I 
have described elsewhere (Quart. Journ. Micr. Sci., vol. xxii., p. 283, plate 35, 
fig. 19). The wings themselves develop in a manner very similar to those of the pinnae 
of Cyccts or Stangeria, until the special peculiarities accompanying the formation of 
the vascular bundles and ribs make their appearance. Comparing the leaf of Gnetum 
