i 84 Lady Isabel Browne. 
that the megasporangia of Selaginellites elongatus contain twenty 
to thirty spores gives some support to the view that this species, 
coming from the Middle Coal Measures (18), is relatively primitive. 
In the strobili of Selaginella the sporangia originate not from the 
sporophylls, but from their axils, or even from a group of cells 
belonging to the axis, situated distinctly above the sporophylls (5). 
In the fossil Miadesmia the sporangia are epiphyllous. This con¬ 
dition is probably primitive, for in all the known palaeozoic Lycopods 
the sporangia seem to have been borne on the leaves. It was con¬ 
cluded in the first and second articles of this series that the 
sporangia of the Sphenophyllales and Equisetales were also 
primitively leaf-borne ; and if, as seems likely, these two phyla had 
a common origin, though a very remote one, with the Lycopodiales 
we should expect the sporangia of the latter to be epiphyllous in 
the more primitive forms. Professor Bower, however, regards the 
epiphyllous or axial position of the sporangium as an overrated 
character (5). 
Miss Lyon has shown that in Selaginella rupestris only one mega¬ 
spore usually matures, though exceptionally there may be two (17). 
In most species there are four megaspores in a sporangium ; a less 
advanced type of heterospory occurs in the Carboniferous Selagi- 
neUites suissei and 5. elongatus where the megaspores seem to have 
been more numerous, there being sixteen to thirty in a single 
sporangium (12). Certain Coal Measure species, however, had, like 
the ordinary species of Selaginella , only four megaspores in each 
sporangium (12), (20). The reduction of the megaspores to a single 
one is clearly a later specialization and connected with it we find, 
both in Miadesmia and in Selaginella rupestris , a retention of the 
megaspore within the sporangium and the formation of a seed. 
Even before the discovery of its sporangia M. Bertrand recognized 
the probable affinity of Miadesmia to Selaginella (3). In its anatomy 
it resembles the more primitive species of Selaginella , but the 
seeds of both genera were, nevertheless, clearly homoplastic 
developments, for those of Miadesmia are morphologically far more 
highly developed, since they are provided with a velum arising from 
the proximal end and leaving a circular micropyle at the distal end. 
Nevertheless the earlier stages of germination do not appear to 
have taken place within the spore of Miadesmia, as they do in the 
comparatively little specialized seed of Selaginella rupestris (16). 
As regards the embryology, Dr. Campbell and others have 
figured the development of the suspensor arising by the first division 
