mer is terns of Ferns as a Phylogenetic Study . 375 
here, the sporangia being of a still more bulky type, and 
massively but shortly stalked, the protection from evaporation 
from their surface when young is less necessary, and the 
indusium is no longer present : though the wall of the 
sporangium when mature is still only one layer of cells in 
thickness, it is more robust in early stages than in the 
proceeding types. I have pointed out (Figs. 77-78, 82) that 
extra oblique segmentations occur in the basal part of the 
sporangium, and that the tapetum may divide into more than 
two layers (Fig. 82), though this is the typical number for the 
simpler Ferns 1 . The number of spores produced from each 
sporangium is also very large 2 : thus the sporangia are re- 
latively fewer, but more bulky and shorter stalked, and produce 
individually a larger number of spores : these are all characters 
leading towards the Marattiaceae. Finally, I have observed 
occasionally in Todea barbara a coalescence of two sporangia 
together. But this which is exceptional in the Osmundaceae 
is the rule in the Marattiaceae ; not only are the sporangia 
sessile, large, and massive, with thick walls, and producing 
very numerous spores, but they are aggregated together in 
confluent masses, while an indusium is absent. These characters 
I regard as further indications of adaptation for protection 
against drying up while young, the lateral coalescence 
especially would well qualify the Marattiaceae for success- 
fully developing their sporangia in exposed situations. 
A comparison of these coalescent sori of the Marattiaceae, 
on the one hand with the sori of Marsilia , and on the other 
with the large and deeply sunk sporangia of Ophioglossum , 
led Strasburger 3 to conclude that the latter are not simple 
sporangia but ‘sporocysts,’ which, resulting from a coalescence 
of sporangia more complete than that of the Marattiaceae, 
really are, according to him, not homologous with the 
sporangia of the true Ferns, but with sori : this view he 
1 Compare Goebel’s Fig. 103 D ; Schenk’s Handbuch, III, p. 388. 
2 Twenty-three spore-mother-cells are seen in the section shown in Fig. 82, while 
thirty are seen in section in Goebel’s figure above quoted. 
3 Bot. Zeit. 1873, p. 84. 
