93 



10 — 20 mm broad. Lacuna? muciferse are present in root fibres forming 

 a continuous ring beneath the cortical layer. 



f. obovata. Stipe shortest, 1-2 cm long, Snbcompressed, with a ring of mueiferous lacuna? • 

 lamina obovate or roundish, broadly oval or somewhat cordate at base, 25-55 cm long, 25-30 

 •cm broad, with undulated margin. 



Hab. Shikotan, Urupp. 



Stipe is shortest, subcom pressed, lamina is shortest and broadest, 

 being obovate, roundish, or somewhat oblanceolate, more or less dilating 

 upward; the base is broadly ovate or somewhat cordate; margin is 

 very much undulated. Remaining characters are same as the other 

 forms above described. 



In instituting a comparison of these three forms we find a gradual 

 transition of form from one to the other. Rugose surface of lamina, 

 presence of lacunas in stem, the median fascia, gyrato-appearance of 

 sori, — all those characters remaining common, the lamina varies from 

 linear to obovate. The length and breadth of lamina of those three 



forms taken in order read as follows : 170 -f cm by 3 — 6 cm, 



50 — 100 by 10 — 17, and 25—55 by 25 — 30. The true length of 

 lamina of f. linearis is unfortunately unknown, the upper portion of 

 specimens in my possession being cut off; and that of others is in 

 some measure insufficient, owing to the erosion of the upper portion. 

 Although the true length of lamina is insufficient, yet the increasing 

 breadth of lamina will tell us the corresponding decrease in length. 

 In company with the variation of form of lamina, the stem varies 

 somewhat in length, the stem of /. latior being longer than that of 

 the rest. From this reasoning, I think that /. latior should be considered 

 the typical or original form of this species, from which the other two 

 forms were derived. 



Root- formation. In the course of my study, I have studied in 

 /. latior the formation of root. Though not quite sure that the study 

 of the same will add some novelty to the branch of science, yet as such 

 a young plant as that may prove profitable to the study of root-formation 

 can rarely be obtained, I should think it better to describe here than 

 to pass it over in silence. Specimens which have been sought for, 

 have been two in number (Fig. 8 — 9). The one has 2 mm long stipe 

 which expands gradually into linear lamina at the upper portion. 



