690 PROFESSOR F. O. BOWER ON 



Such examples as those quoted serve to illustrate steps of transition between open 

 venation and reticulation. The fossil record indicates that the former is relatively 

 primitive. The structure of the juvenile leaves accords with this, the dichotomous 

 venation being clearly the foundation, with the reticulation superposed upon it. 



Hymenoph yllacese. 



This family of Ferns is now generally accepted as a relatively ancient one, but 

 highly specialised in relation to a moist habit ; the filmy structure of its leaves is no 

 longer regarded as an indication of affinity to the Bryophyta, as was formerly 

 believed. Less importance may accordingly be attached now to the details of its 

 leaf-architecture than at the time when Prantl wrote his Monograph. Nevertheless, 

 their importance should not be unduly discounted, as indicating primitive characters 

 of general application, and anatomical comparison with the Zygopteridese strengthens 

 this position. 



The branching and venation of their leaves has been studied very closely, and 

 Prantl in particular has analysed them in relation to dichotomous branching.* 

 He referred all the diverse forms of leaf seen in the family to the development of a 

 dichotomous system. In certain cases the leaf is simple in form, with entire margin, 

 and no apparent branching of the blade. This is traversed by veins, and the general 

 construction is like the "webbed" types in other Ferns. Where the leaf-blade is 

 circular, the whole branching system of veins may show very perfect and equal 

 dichotomy, as in Trichomanes reniforme, and both shanks of each forking develop 

 with an almost diagrammatic regularity (fig. 23). But in certain other entire leaves 

 that equality is not maintained, as in Trichomanes punctatum (fig. 24, a) ; and 

 steps may be found leading by easy stages to a high inequality of the shanks, with 

 the continual assertion of a sympodial false axis (fig. 24, b, c). But if such steps 

 may be followed in the webbed or unbranched types, much more plainly is this to 

 be seen in those which are profusely branched. The analysis of many forms from 

 this point of view was carried out by Prantl."!" In view of his observations it will 

 be unnecessary to enter into further detail here as to the relation of the branching 

 to the cellular construction. He concluded]: that all complicated leaf-forms, such 

 as he described for the Hymenophyllacese, were referable to a simple type of 

 dichotomous branching, and he suggested that a similar method might be applicable 

 also to other families of Ferns. He applied it himself to the Schizseacese, and we 

 now see that its application extends to the juvenile leaves of other primitive 

 types of Ferns. 



Prantl was disposed to regard the fully webbed types, such as T. reniforme, as 

 relatively primitive. This is a position open to doubt, and especially so in view of 

 the facts relating to many other Ferns. It would seem more probable, and it certainly 

 accords equally with the facts, that the entire leaves are a result of " webbing " in an 



* Hymenophyllacee?i, Leipzig, 1875. t L.c, p. 7, pi. ii. J P. 14. 



