3o8 CYATHEACEAE [CH. 



diminution in size, together with a difference in dehiscence. This involves 

 also a change from median to obliquely lateral dehiscence: i.e. from von 

 Goebel's "longicide" to his "brevicide" type. Notwithstanding his doubts 

 as to the reality of such a change {Organographies II. Aufl., p. i i8o), when a 

 series shows so many points of progressive similarity as the Gleicheniaceae- 

 Cyatheaceae sequence does, the probability appears strong that the phyletic 

 change did actually take place. A similar question arises in regard to the 

 marginal series also. A circumstance that will help towards elasticity of view 

 in such matters is the existence of "looking-glass" images of sporangia from 

 the same plant, such as are seen in Plagiogyria (Fig. 546), Loxsovwpsis (Fig. 

 524), Th}'rsopteris{¥\g.^2g,E,F), and others. These show that in a "brevicide" 

 type the stomium may lie on either the right or the left side of the ring. If 

 this difference can exist between sporangia from the same individual plant, 

 it shows that the position of the stomium is not a fixed feature. If it can lie 

 either right or left in the closed ring, why not either in a median or in 

 a lateral position? Further, in the organisation of the stomium there is 

 distinct advance within the series from Gleiclienia, through Lopliosoria, to 

 the Cyatheaceae, and this goes with the diminution in size and in the spore- 

 output. 



As we progress from the Gleicheniaceae to the Cyatheaceae there also 

 follows a change in the segmentation of the young sporangium, which is 

 probably connected with the proportion of the sporangial head to the length 

 and thickness of the stalk. A two-rowed cleavage is constant for Metaxya, 

 and for the Cyatheaceae generally. The constancy of this peculiarity here, 

 and its absence (so far as observed) in the Dicksonieae, is a distinctive feature 

 for the Family, which it shares with the Dipterids only, so far as present 

 observation extends. This gives it a special diagnostic value. The spore- 

 numbers have shown the Cyatheaceae to have already fallen to the figure per 

 sporangium usual for Leptosporangiate Ferns, which is so much lower than 

 in the Gleicheniaceae: but a climax is reached in C. dcalbata, where the 

 numbers may be unusually small. Comparing these results with the numbers 

 of spermatocytes seen in vertical section of the antheridia, their relation 

 comes out similar to that seen in such Ferns as Djjopteris filix-nias (see 

 table, Vol. I, p. 292). 



The result of such comparison of the series of Ferns with superficial sori, 

 in respect of these various criteria, is to show with impressive uniformity a 

 progression from the ancient type of the Gleicheniaceae to the more modern 

 though still relatively primitive Family of the Cyatheaceae. In all of them 

 the position of the sorus is consistently superficial. As the Gleicheniaceous 

 type is of proved antiquity, and possibly even of Palaeozoic origin, it appears 

 that Ferns with a superficial sorus, linked with the Cyatheaceae by living 

 genera showing intermediate characters, have had a consecutive history as 



