Lycopodiales. 185 
of the fertilized ovum (10). Miss Lyon, however, found none in 
Selagiiiella apus (18) ; probably its absence there is a secondary 
character, for its absence is hitherto only recorded in that species, 
and a suspensor, though one of varying importance, is characteristic 
of Lycopodium, which, though not closely related to Selaginella, 
probably comes nearer to that genus than any other living form. 
As regards the affinities of the Selaginellacese with the orders 
already discussed, they show many points of resemblance to the 
Lepidodendraceae. There is a distinct resemblance between the 
general anatomy of the exarch protostele of the erect stem of 
Selaginella spinosa and a protostelic Lepidodendron. This 
probably does not indicate any direct connection between the two 
orders, for an exarch protostele is found in what are supposed to be 
the more primitive members of all the orders of the Lycopodiales. 
Other less important resemblances are the Stigmarian appendages, 
comparable to the monarch roots of Selaginella, and the secondary 
growth in thickness of Selaginella spinulosa. This secondary 
growth is, of course, a freshly acquired character not found in the 
other species ; the mode of the secondary growth, moreover, is 
different from that of the Lepidodendraceze and the analogy 
indicates no especially close affinity. The ligule, on the other hand, 
indicates a real, though not a very close, affinity. Selaginella 
appears to be a genus showing several modes of increase in com* 
plexity, but the complications all seem to be derivable from an 
exarch protostelic form. There is therefore no reason to suppose that 
the genus arose by reduction from the Lepidodendraceag. Moreover 
we have evidence proving the existence of heterosporous herbaceous 
Palaeozoic Lycopods and it is therefore very probable, as maintained 
by Dr. Scott, that Selaginella was derived from such herbaceous 
forms (20). In any case of the seeds of Miadesmia, Selaginella and 
Lepidocarpon each forms a culminating point of a distinct evolu* 
tionary series : that these genera never had a common seed-bearing 
ancestor is clear from the totally different structure of their seeds. 
The relationship of Selaginella to Isoetes is neither close nor 
clear; the absence of a strobilus or a suspensor in the latter, its 
relatively large leaves and epiphyllous sporangia, are all in striking 
constrast to Selaginella. The last character, however, is also found 
in Miadesmia, a genus that probably belongs to the Selaginellaceous 
cycle of affinity. As it is one of the oldest known members of that 
cycle it may well have been primitive in that respect. As pointed 
out in considering the affinities of Isoetes to the Lepidodendraceaa 
