514 FILICALES 
suggested, and structurally the sorus itself offers no difficulty (compare 
Figs. 283 H and x). It is thus seen that the general type of sorus is 
constant in the family: the chief differences lie in the mode of associa- 
tion of the sporangia, and in the extent and fission of the sori. 
As the development of the individual sporangium has been found to 
be essentially the same in the several genera, notwithstanding the difference 
that exists between the synangial state and that where the sporangia are 
separate, it will suffice to describe it for one only, and Angéopterts, with 
which Archangiopteris shows close similarity, may be selected as being 
the most familiar! At an early stage the sporangia begin to project as 
separate upgrowths; but it is impossible, from a study of superficial sections, 
to detect any regular system of segmentation which is maintained in all 
sporangia; a comparison of the four sporangia, shown in Fig. 284 a in 
surface view from above, discloses no regular sequence of segmentations, 
and the cell-groups which’ will develop into the sporangia appear con- 
sequently ill-defined. Of the sporangia a, 4, c, d, shown, that marked 
() is believed to be the most regular and usual type; and the cells 
shaded in it are evidently sister cells, derived from a single parent cell 
which, as we shall see, gives rise to the central and essential part of 
the sporangium; we may call this, as in the other genera, the superficial 
parent cell. If a section were taken along a line x—. through such a 
sporangium, after it had grown more convex, it would appear as Fig. 284 B, 
in which the cells shaded are believed to correspond to those shaded 
in Fig. 284 a. It becomes apparent from such sections as these that 
a single cell, the central cell, had divided periclinally to form an inner 
cell and a superficial one; the former is the archesporium, and has in 
Fig. 284 B already divided into two; the superficial cell has also divided 
repeatedly. Though we may thus select sections so as to represent a 
reasonably regular arfd typical structure of the young sporangium, it is 
clear, from Fig. 284 a, that what has been described is only a central 
type, and as a matter of fact hardly any two sporangia show exactly the 
same details of segmentation. As development proceeds, growth and 
cell-division often continue with sufficient regularity to allow the genetic 
grouping of the tissues to be clearly followed (Fig. 284 c). Meanwhile, 
certain cells at the apex enlarge to form the crest-like annulus. The 
relation of this to the main lines limiting the product of the superficial 
parent cell is variable; a common case is that shown in the figure, where 
the middle line (x) coincides with the limit of the annulus; but this is 
by no means constant: from this point (x) downwards, on the central 
side of the sporangium, the dehiscence will take place. The sporogenous 
group is now clearly defined by the character of its protoplasmic body, 
and it forms a definite block of cells, referable to a single parent. Next 
follows the change of the cells immediately surrounding the sporogenous 
1Compare ‘‘ Studies,” iii., P27. Trans, 1897, where more full details are given for 
this and other genera. 
