Anatomy of the Gemis Selaginella , Spr, 23 
In group D the two kinds of sporangia are approximately of the same 
size, neither showing any great degree of complexity. It is of interest 
to note that the species in this group belong almost exclusively to the 
6 *. Martcnsii group ( 13 ), and that the indefinite character of the cone 
is associated with a corresponding simplicity of stem structure ; otherwise 
the classification of the species as based on the anatomy is not followed 
definitely by peculiarities of sporangial arrangement. 
The occurrence of imperfect sporangia. 
Bower ( 3 ) states that ‘ no cases of alternating sterile and fertile zones 
are recorded, but in some species the fertile spikes revert at the apex 
to a vegetative character. Abortive sporangia have been seen at the base of 
the strobilus of .S', spinosa and 5 . Martensii , and would doubtless be found in 
many species, but no isolated sporangia have been seen in the sterile region.’ 
The occurrence of abortive sporangia in Selaginella is a far more wide¬ 
spread phenomenon than one would be led to suspect by Professor Bower’s 
statement. In most cones, scattered but imperfect sporangia are to be 
found. In Kraussiana , the majority of the sporangia never reach 
maturity, but are arrested at the stage figured in PL IV, Fig. 12, the whole 
being of a distorted shape, consisting of a two-layered wall, stalk, dis¬ 
organized tapetum, and sporogeneous cell-mass in the interior. Those 
which do develop are large, massive structures, but they are few in number. 
In S. Helvetica , V. Wallichii , V. or eg ana, and S.flabellata, all of which 
produce elongated cones, the sporophylls of the three latter being par¬ 
ticularly crowded and numerous, the whole of the middle region is 
frequently sterile (cf. Lycopodium Selago). 
At the base of every cone in involvens , the megasporangia are 
persistently abortive. The mature megasporangia are found above this 
region, and are large, and contain four dark, thick-walled spores. The 
same is true also of 5 . gracilis , where the basal sporangia are invariably 
shrivelled. vS\ viridangula progresses a step further. Here, the basal 
sporangia are entirely abortive, and the basal sporophylls are purely 
vegetative in function. 
That the definite sterilizations quoted above (whether occurring in the 
middle, or at the base, of the strobilus) are utilitarian, and make for the 
increased efficiency of the spore production as a whole, is well illustrated 
by the case of S. Kraussiana. This species is fairly delicate, and the 
strobili are developed in profusion, in clusters at the tips of the branches. 
The supply of nourishment is insufficient to bring all the sporangia to 
maturity. Most of them, therefore, abort at an early stage in development, 
and the subtending leaves become purely vegetative. These sterilizations 
preserve the balance between the spore-producing and the vegetative parts, 
and are best explained as a continuation of the process by which the 
