358 Bower . — The comparative examination of the 
the following results : the apex of the young pinnule appears 
rounded, so that transverse sections of it are circular, the 
formation of the wings not having begun ; but lower down the 
wings appear as a rounded weal on either side : here (Fig. 70, x) 
the superficial cells increase in size, and are more densely 
filled with protoplasm than the rest: the projecting wing which 
thus originates does not spring from a single row of cells ; 
the transverse section clearly shows that a number of cells 
take part in its initial formation, and these continue to grow 
and divide, while the rest of the pinnule, which finally con- 
stitutes the midrib, passes over into the quiescent condition. 
The cells forming the wing are of rectangular form (Fig. 70) 
and undergo repeated T divisions (Figs. 71, 72, 73); pyra- 
midal cells are absent with the exception of those which form 
the hairs. As to the exact sequence of segmentation in the 
growing wing, I have not been able, any more than in Todea 
barbara , to recognise any invariable rule : sometimes (Figs. 71, 
73) a group of cells, having obviously a common origin, 
occupy the centre of the transverse section, and in such cases 
it might be possible to recognise a certain cell as the perma- 
nent initial : in other cases (Fig. 72) the wing obviously 
originated from the outgrowth and repeated divisions of 
several cells, and the whole section of the wing may be seen 
to be composed of a number of groups of cells, each having 
apparently a distinct origin. I conclude that here, as in Todea 
barbara , though T division is the universal rule, there is no 
definite arrangement of the groups of cells composing each 
transverse section, so as to indicate their origin from any 
single initial. This mode of growth and segmentation of the 
wing is maintained throughout its development 1 , the result 
being a massive structure of the wing, consisting of ten or 
more layers of cells. 
The results of the above observations on the characters and 
mode of development of the wings of the leaves of Ferns may 
now be drawn together, and compared : the following are the 
chief points to be noted : — 
1 Phil. Trans. 1884, Part II, Plate 38, Fig. 19. 
