loo DRYOPTEROID FERNS [CH. 



of the sporangia is basipetal. The habit and the characters of the sorus are 

 in fact those of the smaller Cyatheaceae, of which the genus Woodsia may 

 be regarded as including arctic and mountain representatives. 



Fortunately the prothalli and sexual organs are particularly well known, 

 and the details confirm this suggestion. The prothalli were found by Schlum- 

 berger to be highly variable under special conditions of culture {Flora, 191 1, 

 p. 384): particularly under weak illumination they may take a filamentous 

 form (Vol. I, Fig. 266). Accordingly he fixed upon features other than mere 

 form that appeared of greater value for comparison, viz. the occurrence of 

 certain hairs, and the structure of the antheridium. Bristle-shaped pluri- 

 cellular hairs on the surface and margin of the prothallus were held by Heim 

 to be characteristic for the Cyatheaceae {Flora, 1896, p. 355, etc.). Hairs of 

 like form, with or without a glandular terminal cell, were found in W. obtiisata, 

 and also in Diacalpe aspidioides, thus providing a common feature. More 

 cogent evidence, however, may be drawn from the structure of the antheridia, 

 and particularly from the behaviour of the lid-cells (/.r.p. 387). It has been seen 

 in Vol. I (p. 293, Fig. 283) how, with the diminishing number of the sperma- 

 tocytes seen in passing from the Eusporangiate to the Leptosporangiate 

 Ferns, there is a simplification of structure of the antheridial wall. In 

 particular this appears in the divisions of the cap-cell. In the Cyatheaceae 

 this is as a rule divided into two or more cells, while in the "Polypodiaceae" 

 it is undivided. Now in Diacalpe it is found that the division is as in the 

 Cyatheaceae, viz. usually into two, but occasionally into three cells. But 



Woodsia obtiisa shows only two cells, and occasionally the cap- cell remains 

 undivided. In W. ilvensis this undivided state is the rule. Schlumberger states 

 that Diacalpe and Woodsia differ from all the "Polypodiaceae" in having 

 a divided cap-cell : and Von Goebel adopts this so far as investigation has 

 extended {Organogi^aphie, II, p. 925). The rarity of such exceptions together 

 with their comparative bearings confer a special interest upon these facts 

 relating to the antheridia of Diacalpe and Woodsia. 



The interest extends equally to the sporophyte, and in particular to the 

 sorus. The axis in Woodsia is short, and ascending or upright. Both stem 

 and leaves bear superficial scales, which are specially prominent in W.poly- 

 stichoides. The character of the relatively simple leaves of the small mountain 

 species, together with the sori which they bear, is seen in Fig. 646, A-E, but 

 in certain larger American species such as W. mollis (Klf.) J. Sm., obtiisa 

 (^Spreng.) Torr, scopuliiia Eat., and mexicana Fee the leaves may be doubly 

 pinnate, recalling the more elaborate Cyatheaceae in habit. Following on a 

 protostelic and solenostelic stage in the young plant, the vascular system of 



Woodsia widens out upwards into a simple dictyostele. From near the base 

 of each foliar gap arise two separate leaf-trace strands, which enter the 

 petiole, and beneath their point of origin passes out the supply to one or 



