22 Lady Isabel Browne. 
facts, and the circumstance that these phenomena are confined to 
the two anatomically most complex genera, suggest that the 
secondary tissues have been recently acquired. 
Professor Bower regards the sorus of free sporangia, found in 
Archangioptens and Angiopteris, as a modification of the synangial 
sorus found in the three other genera. The priority of one or other 
of these types is, as he asserts, not clearly proved ; further, the 
older fronds with non-soral sporangia cannot, as he points out, be 
used as evidence of the phylogenetic development within the sorus 
itself (5). But against the primitiveness of the synangium among 
the Ferns it should be urged that the sporangia of some Pterido- 
sperms are more or less coherent, and as the latter are a relatively 
ancient group, some of the older synangia probably belonged to 
them. Mr. Kidston has shown that the cohesion of the sporangia 
in the Pteridosperms was probably secondary (27), and this favours 
the view that the common ancestor of Ferns and Pteridosperms 
possessed free sporangia, especially as the sporangia of the oldest 
distinctly cryptogamic fern-like plants, the Botryopterideae, were 
typically (though not invariably) free. It has often been assumed 
that since some Psaronieae bore Pecopterid fronds, and since some 
species of Pecopteris bore synangia, these synangia belonged to 
Psaronius ; this supposition has been held to favour the primitiveness 
of the synangium in the Marattiacese. But as at least one Peco¬ 
pteris was a Pteridosperm, and as its microsporangia are unknown, 
some or all of these synangia may belong to the Pteridospermeae 
(18). As we are not certain that the fructifications of the Psaronieae 
were synangial, or even that Psaronius was a fern, this argument is 
of little value. 
But Professor Bower relies chiefly on the developmental 
evidence to show that the synangium is a sporangium multilocular 
by septation ; if so, as the free sporangia of Angiopteris are clearly 
homologous with the loculi of the synangium of Marattia, the sorus 
with free sporangia is a derivative condition (5). In support of the 
hypothesis that the septa between the loculi are sterilized sporo- 
genous tissue, Professor Bower urges that though the archesporium 
usually arises from a single cell, yet a cell-packet outside the 
definitely fertile group may become sporogenous, or conversely 
certain cells usually sporogenous develop as tapetum. The latter 
case must be admitted to be an example of sterilization, such as 
occurs in the sporangia of many Pteridophyta. That the arche¬ 
sporium of each loculus is not always referable to the division of a 
single cell can hardly, however, be held to indicate that these loculi 
