MARATTIACEAE 519 
have been noted also in other genera, so that Danaea does not stand 
alone, though the demonstration is most effective in that genus. 
In putting a morphological interpretation on these facts it is to be 
remembered that the genus Danaea has in many of its species very 
long sori compared with those of MJarattia. External comparison had 
already suggested an elongation of the type of the latter genus to produce 
the sorus of Danaea. The internal structure is now seen to be conform- 
able to such a progression, for the partial septations are commonly 
found in those sporangia which are larger than the normal: they thus 
appear to have followed upon expansion of the sporangia. Their existence 
Pecopteris (Dactylotheca) dentata, Part of a frond with the Matrix. Reduced after a 
photograph by Mr. W. Hemingway. From Scott’s Studies in Fossil Botany. 
here, as well as occasionally in other genera, raises the question whether 
Or not septation has been effective in the primary production of the 
sorus: this will be considered again later. 
‘The synangial sori differ from those with separate sporangia in the 
mechanism which accompanies dehiscence, though the dehiscence itself is 
uniformly by a radial slit. The annulus represented in Amgiopteris by a 
broad indurated band or hoop, is absent: clearly it would be useless in 
a synangium, for it could not possibly be effective unless the sporangium 
were free from mutual relations with others. In KXawdfussia and Danaea 
the radial slit of each sporangium may widen as the neighbouring cells 
dry up, to form an almost circular pore. In Afarattia there is in addition 
to this a change of form of the whole sorus at maturity: its two sides, 
originally in close juxtaposition so that the distal ends of their sporangia 
almost touched, move apart like the opening of a book. In Danaea 
and Marattva the whole external wall of the synangium is composed of 
deep indurated cells: the condition of Axgiopteris appears accordingly 
