56o MOSSES AND FERNS chap. 



tic perispore. Within the special membrane is d'*iveloped a sec- 

 ond membrane — exospore — which later shows a division into 

 three layers. Within the exospore the mesospore and endo- 

 spore arise very much as in Selaginella, which Isoetes further 

 resembles in the separation of the mesospore from the protoplast 

 and from the exospore, although this is less conspicuous than 

 in SclagincUa. 



As the sporangium develops, the surrounding leaf tissue 

 grows up about it, somewhat as the integument of an ovule 

 invests the nucellus. Goebel calls attention to the resemblance 

 between the sporangium of Isoctcs, sunk in the fovea and par- 

 tially covered by the velum, and an ovule with a single integu- 

 ment. 



Bower finds in the sporangium of Lepidodendron, structures 

 which resemble the trabeculae of Isoetes, and he is inclined to 

 consider the two genera as really related. 



In /. lacustris the sporangium is sometimes replaced by a 

 leafy bud which may develop into a perfect plant. (Goebel: 

 "Ueber Sprossbildung aus Isoetesblatter," Bot. Zeit., 1879). 



The relationship of Isoetes to the other Pteridophytes is not 

 entirely clear, and there has been a good deal of difference of 

 opinion on this point. In many respects it shows a nearer 

 affinity to the eusporangiate Ferns, than to the Lycopodinece, 

 in which the genus is usually included. The archegonium 

 closely resembles that of Ophioglossum or Marattia, and the 

 spermatozoids are multiciliate, which is never the case in any 

 known Lycopod, but is universal among the Ferns. The 

 anatomy of the sporophyte is quite peculiar, but may, perhaps 

 be quite as aptly compared to the Fern-type, as to that of the 

 Lycopodineae. The dichotomous branching of the roots has a 

 parallel in Ophioglossum, although it must be admitted that it 

 closely resembles the forking of the root in Lycopodiiim. The 

 sporangium may perhaps as well be compared to the spike of 

 Ophioglossum or the synangium of Dancca as to the single 

 sporangium of Lycopodiiim or Lepidodendron. It would be 

 rash to assert positively that the trabeculse correspond to the 

 partitions between the sporangia of Ophioglossum, and that 

 the sporangium is really compound, but this is not inconceivable. 

 The position and origin of the large sporangium of Isoetes are 

 certainly not very unlike those of the sporangiophore of 

 OpJiioglossum. 



