XVII SUMMARY AND CONCLUSIONS 597 



ably all secondary, the Pteridophytes show four types of gameto- 

 phyte. The first, represented by most homosporous Ferns, is 

 the familiar heart-shaped prothallium, which strongly recalls 

 the simpler anacrogynous Jungermanniaceae or Dendroceros; 

 the second is the lobed prothallium of EquiscHim, which resem- 

 bles most nearly among the Liverworts such forms as Antho- 

 ceros fiisifonnis, but has an analogy also in the lobed prothallia 

 sometimes met with in Osmiinda. In some species of Trich- 

 omancs and Schiz(Ta there occur the branched filamentous pro- 

 thallia, which some authors look upon as an indication of direct 

 relationship with forms intermediate between Algse and Musci- 

 neae. As other species of Trichomanes have the same type of 

 prothallium as the. other Ferns, and this is always true of the 

 closely related genus HyinenophyUnm, this view is open to 

 question. The green prothallium of Lycopodhun ccrmmm dif- 

 fers from the somewhat similar one of Equisehim, in the essen- 

 tial point that in the former we have to do with a radial 

 structure, in the latter with a dorsiventral one. The upright 

 gametophyte of Lycopodiiim, with its terminal circle of leaf-like 

 lobes, might be compared to a leafy Moss-shoot, although it is 

 hardly probable that this resemblance is more than superficial. 



As far as the form and growth of the prothallium are con- 

 cerned, all forms except Lycopodium could be traced back to 

 the Anthocerotes ; the Fern type to forms like Dendroceros or 

 Anthoceros Iccvis, the Equisetnm type more resembling A. fiisi- 

 formis. The difference in the character of the chromatophores 

 is a very important one, and at present must forbid the assump- 

 tion of any immediate connection between the Anthocerotes 

 and existing Pteridophytes. Whether the occasional appear- 

 ance of very large plate-like chromatophores in the prothallium 

 of Osmiinda cinnamomea (Campbell (12)) is a reversion to a 

 primitive condition retained in the Anthocerotes, it is, of course, 

 impossible to say, but it is not" inconceivable, especially as the 

 same thing is found again normally in the sporophyte of Sel- 

 aginella. The regular doubling of the chromatophores in the 

 sporophyte of Anthoceros also suggests that the multiple chro- 

 matophores of most Archegoniates are secondary. 



In the Anthocerotes the origin of the archesporium is differ- 



^ent from that of the other Hepaticse, being hypodermal, as in 



the lower Pteridophytes. The columella is in position similar 



to the primary vascular bundles in the embryo of the Pterido- 



