A Unit of Construction in the Pteridophyta. 147 
intact and the sporangia are simultaneous in origin, but in Diptevis 
conjugata all stages of the disintegrating process can be seen side 
by side on the same frond. 
Again in Osmunda we see an early phase of the loss of soral 
identity in that the sporangia 'of a single tuft face in different 
directions. A later stage is seen in the scattered sporangia of Todea. 
Such an explanation might conceivably be applicable to 
Schizaeaceee, but Prantl’s view that the sorus has become mono- 
sporangic is far more probable. That reduction of this sort takes 
place is shown in various degrees among the species of Gleichenia, 
and can be traced ontogenetically in the megasporangial sorus of 
Azolla. 
In the case of the Ophioglossaceae the “ sporange ” appears on 
the other hand to represent a reduced sporangiophore comparable 
with the monosporangic form found in Psilotum. It is not un¬ 
common to find sporangia on the so-called sterile segment of Botry- 
chium lunavia. I have also a specimen of B. lamiginosum showing 
the same phenomenon. 
In a specimen of B. lunavia gathered this season in Yorkshire, 
and sent me by Miss Beard, B.Sc., the sporangia are rather on the 
abaxial surface of the margin of the pinnae. Some of them are 
stalked and curve over on to the upper surface, and thus bear a 
curious resemblance to the sporangiophores of Sphenophyllum 
Dawsoni. The dehiscence of each sporange is effected along a 
stomium resembling that of the Psilotaceae. In Ophioglossum the 
fertile spike appears to be a condensed structure with the sporangia 
sunk in an unbranched lobe of the frond. The existence of such 
forms as Mavsilia with fairly typical Fern sori and dorsiventral 
segmentation of the frond seems to justify us in looking upon 
dorsiventral segmentation of the frond in Ophioglossaceaa as a 
character of no great taxonomic value, and as affording no justifi¬ 
cation for the view that the whole fertile lobe is homologous with 
the sporangiophore of the Sphenophyllales. I would also strongly 
deprecate the use of this term in a non-morphological or non¬ 
committal sense, a use which is sometimes adopted by Professor 
Bower in his recent work. 
Evidence has gradually been accumulating during the last few 
years of the convergence of the various groups of Pteridophyta in 
past epochs. Pseudobovnia ursina, a most interesting Devonian 
plant of probably Calamitean affinity, seems to be a type in which 
the microphyllous habit of the Lycopsida had not yet been evolved. 
