Bower. —Sht dies in the Phytogeny of the Filicales. VI. 29 
single cells grow out ( b), enlarge (c), and undergo the usual segmentation. 
Transverse sections through the young sorus show that the stalk is com- 
posed of three rows of cells ; but as one of these does not always extend to 
the base of the stalk, there often appear to be only two (d). As regards the 
distribution of the sporangia relatively to the veins, it is shown in tangen- 
tial sections of the young curved leaf (Text-fig. 15, e) that their distribution 
is very uniform and bears no relation to the veins. By later addition of 
younger sporangia to those first formed, the sorus becomes a ‘ mixed ’ one, 
the later sporangia arising commonly between the veins. The spores 
have one longitudinal angle and two flattened sides, with prickly epispore, 
as against the tetrahedral smooth spores of Syngramme and Metaxya. 
The sum of these characters is consistent with the suggested relation of 
the genera named. The conclusion which may be drawn is that they form 
a phyletic series, of which Metaxya holds the most primitive position, 
Syngramme an intermediate place, and Elaphoglossum is the most advanced. 
They illustrate progress in the substitution of scales for hairs ; in the 
subdivision of the leaf-trace ; in the increasing fusion of the veins of the 
lamina, while the blade itself is simplified in outline ; in the extension and 
fusion of the plural sori on the veins in Metaxya into the elongated sori of 
Syngramme , and the spread of the sori over the region between the veins, as 
seen in Elaphoglossum ; and finally in the introduction of the mixed sorus 
in the last-named genus. All of these progressions are such as are seen 
in other distinct phyla. They are shown in the Dipteroideae. But as they 
occur here in genera in which, there is a consistent absence of vascular twigs 
ending freely in the areolae (a character which is constant for the Dipteroi- 
deae), these genera may rightly be held as constituting a phyletic progression 
of Metaxya, i. e. derivatives parallel with, but distinct from them , to be 
designated the Metaxyoidae . 
The Systematic Treatment of Acrostichoid Ferns. 
The facts and the reasoning contained in the preceding pages have 
demonstrated with a reasonable probability two distinct progressions of 
Ferns, which have led from those with circumscribed sori to Ferns 
with the ‘ Acrostichoid ’ condition. It will be well next to consider more 
generally the various origins of that condition which we are now able to 
trace. As a consequence of the phyletic grouping of Ferns (based not on 
this one character of the sorus, but on the sum of many characters) there is 
reason to believe that they are numerous. Sir William Hooker, using the 
features of the sorus only, adopted an extreme position . 1 His genus 
Acrostichum included very diverse types. In explanation he remarks : 
‘ Many have been the attempts to divide the species into a number 
of distinct genera, but, as will be seen by the synonyms I have quoted, not 
1 Species Filicum, vol. v, p. 194, 1864. 
