XXIX] DICKSONIEAE 267 



with connective cells of the ring that are not indurated. But the number of 

 the cells of the annulus and of its parts does not appear to be constant. 

 The orientation of the sporangia is not uniform, especially at the margins 

 of the flattened receptacle: but the majority have the distal face turned 

 outwards from the receptacle. The output of spores is not large: the typical 

 number is 64. 



Hitherto the prothalli of the Dicksonieae appear to be imperfectly known. 

 But recently prothalli of Cibotium Barometz and of Dicksonia antarctica 

 have been raised in pure culture by Prof. A. G. Stokey of Holyoke College, 

 Mass., U.S.A. She will shortly publish her results: but she permits me to 

 state that the former species, after two years' culture, shows no bristles 

 ("borsten") : in the latter no hairs or scales (excepting rhizoids) appear even 

 on old and branched prothalli. Heim. (/^/c?ra, 1896, p. 362) states, however, 

 that the prothalli and antheridia of the Dicksonieae resemble those of the 

 Cyatheaceae, and bear on margin and surfaces glandular hairs transitional 

 to those of Polypodiaceae. Von Goebel describes bristle-like cell-plates 

 ("borstenformige Zellflachen") on the margin of prothalli of Balantiiim 

 antarcticiim {Organographies II Aufl., ii, p. 950). These statements do not 

 accord with the results of Prof. Stokey, and the matter awaits further 

 detailed demonstration. 



Both structurally and in the characters of the sorus and sporangium the 

 Dicksonieae, in the restricted sense, occupy a position related closely to 

 Thyrsopteris, and less directly to the Loxsomaceae and Hymenophyllaceae. 

 But turning from these relatively primitive types they show also a relationship 

 to the Dennstaedtiinae, so close indeed that Hooker included Dennstaedtia 

 as a sub-genus of Dicksonia, under the name of ^Patania Presl. The signs 

 of advance which these Ferns show as compared with the Dicksonieae fully 

 justifies their separation, with Microlepia and some others, as a natural sub- 

 tribe. It will be seen that certain definite characters of advance in their sori 

 uphold this treatment. 



It seems probable that some of the Jurassic fossils grouped under the 

 name Coniopteris may have really been early Dicksonioid types. 



BIBLIOGRAPHY FOR CHAPTER XXIX 



Hooker & Bauer. Genera Filicum. Tab. 44, A, 1842. 



529. Hooker. Species Filicum. i, p. 64, 1846. 



530. KuNZE. Farnkrauter. Leipzig. 1847. 



531. Engler & Prantl. Natlirl. Pflanzenfam. i, 4, p. 122. 



532. Bower. Studies in Spore-producing Members. IV. Phil. Trans. Vol. 192, p. 67. 



533. Bower. Origin of a Land Flora. 1908, p. 588. 



534. Bower. Studies in Phylogeny. III. Ann. of Bot. xxvii, p. 454, 1913- 



535. Seward. Fossil Plants, ii, p. 367, 1910. 



520 



