36 DAVALLIOID FERNS [ch. 



acquired from their presumable Dennstaedtioid ancestry are recognisable 

 by comparison. They mark the transition from a gradate to a fully Lepto- 

 sporangiate state, but with the individuality of the sorus retained. 



There are, however, certain further details of the sori that claim special 

 attention : the vegetative characters may be held as subsidiary. Of the three 

 modifications specially noted in the sori of Davallioid Ferns, the first is the 

 passage of the receptacle from the originally marginal position of the 

 Schizaeoid and Dicksonioid Ferns to the superficial position, which is more or 

 less marked in the Davallioids. This may properly be held as a change of 

 advance. The gradual steps are illustrated by such a series as Thyrsopteris 

 (Vol. II, Fig. 529), Cibotium (Vol. Il, Figs. 534, 535), Davallia (Fig. 590), 

 Lindsaya (Fig. 6o'^,d.x\A]>! cpJirolepis (Fig. 593). Until the fuller developmental 

 details are to hand for the last of these, it must remain uncertain whether the 

 receptacle itself has ever in this group of Ferns slid from the margin to the 

 surface in point of initiation. In all the other examples the receptacle is 

 actually marginal in origin: but in all of them, in more or less degree, the 

 sorus shifts towards the lower surface in course of the individual development, 

 and may, when fully matured, appear as though seated far from the margin 

 (compare Figs. 588, 592). As in Hypolepis the • upper indusium may be 

 assimilated to the general expanse of the blade, and this is a considerable 

 factor in producing the final result. 



The developmental facts in these Ferns disclose a remarkable tenacity in 

 retaining the marginal position for the receptacle. It will be found that in 

 the Pteroids there is a like tenacity up to a point, but that in Pteris itself the 

 sorus has definitely passed from the margin to the surface of the blade: and 

 that this is so not only in the course of individual development, but also in 

 point of initiation (Chapter xxxviii). 



The second change, viz. the elimination of the lower indusium, is already 

 prefigured in Hypolepis (Figs. 585, 586, pp. 10-12). There can be little doubt, 

 in presence of the vestigial remains of it in certain specimens oi Arthroptcris, 

 that there also the inner indusium is subject to reduction or even to complete 

 abortion. The question seems not so clear in Monachosoriim: there the 

 argument is comparative, without the aid of vestigial evidence. There is thus 

 reason to believe that the lower indusium may be eliminated in Davallioid 

 Ferns, as it is in Hypolepis, and as it will be seen later to be eliminated in 

 the Pteroids. 



A third change is the linkage of sori laterally to form coenosori. This is 

 a very natural consequence of condensation from a more highly branched 

 leaf-structure. The highly pinnate leaves of Dicksonioid Ferns retain the 

 individuality of their sori: but in various Davallioid Ferns, and particularly 

 in Lindsaya and Dictyoxiphiuni, the leaf-structure suggests condensation, and 

 coenosori are a marked feature. A vascular commissure accompanies the 



