22 



DAVALLIOID FERNS 



[CH. 



The sori of Nephrolepis are very variable in position, being sometimes 

 marginal with almost equal indusial lips {N. davallioides (Sw.) Kze.), but 

 often also they may be deeply intra-marginal {N. cordifolia (L.) Presl) : in 

 either case they are seated on the vein endings 

 (Fig. 592, c, k). In the latter species they 

 resemble in outline those of Nephrodhnn, but 

 differ in being terminal on the vein (Vol. I, 

 Fig. 224, C, D). So variable a relation of the 

 receptacle to the margin in the adult state 

 raises the question of that relation in point 

 of development. Isolated observations have 

 been made on N. biserrata (Sw.) Schott, in 

 which the sori are distant from the margin : 

 these suggest that the upper indusial lip, which 

 in Davallia has been seen to grow by inter- 

 calary activity (Fig. 590), is in Nephrolepis 

 gifted with a still more active marginal seg- 

 mentation (Fig. 593). The development of 

 the sorus in this genus requires a careful 

 comparative and developmental study: but 

 provisionally it may be held as probable that Fig- 593- A vertical section through 



^ . . the young sorus of l\i ephrol''pis otser- 



all the species were derived from a source rata (Sw.) Schott, showing the great 

 with marginal sori (Studies III, Ann. of Bet. Z^ S :^:^t^^,^. 



191 3, p. 462). the upper (ti) appears very definitely as 



T , . . r \T , 7 1 , • l^ a continuation of the leaf-surface, and 



In Certam species of iV^//zw/^/ZJ there may has an active marginal segmentation. 



also be a lateral fusion of marginal sori, pro- (>< 25°-) 



ducing continuous coenosori, as in Lindsaya or Pteris: this is seen, for 

 instance, in N. dicksonioides Christ (Fig. 592, n-q)^ and N. aaitifolia (Desv.) 

 Christ (Fig. 592, r, s^y). Such facts indicate the high variability within the 

 genus which clearly follows from Christ's striking observations. But notwith- 

 standing such differences of detail he rightly remarks that the characteristic 

 habit gives the true guiding line for the recognition of this remarkable genus 

 {Farnkrduter, p. 288). 



These observations, which might be greatly extended in detail, must 

 suffice to convey what every student of the Davallioid Ferns knows, that the 

 position of the sorus relatively to the margin is very variable. Assuming 

 that systematists are right in their reference of Nephrolepis to a Davallioid 

 affinity, and that it is itself a coherent genus, its species exhibit perhaps 

 more impressively than any other how diverse the relation of the sorus to 

 the leaf-margin may be in nearly related Ferns. The facts may even suggest 

 the facile conclusion that the relation of the sorus to the margin is a character 

 systematically worthless. But the study of soral development in the Dick- 



