XXXIII] COMPARISON 



307 



system is also added in some of the large stems of the Fam.iy. These 

 peculiarly consecutive facts lead by gradual steps from the primitive state 

 of the Gleicheniaceae to the highly complex vegetative state of the Cyathe- 

 aceae: but always with the same underlying scheme. 



A similar progression though not always so consistent characterises the 

 dermal appendages. In the Dicraiioptcris and Eu-Gleicheiiia sections of 

 Gleichenia both hairs and scales are present upon the rhizomes and leaves, 

 developed perhaps in relation to a xeroph}'tic habit. But in G. linearis and 

 pectinata, now separated from the rest of the genus z.?,\Eii-Dicraiiopteris, no 

 scales are found, though in the latter species branched bristles appear at the 

 leaf-bases, borne upon massive emergences. In Metaxya and Lophosoria hairs 

 only are present; but in the Cyatheaceae broad scales cover the stem and young 

 leaves, while at the leaf-bases they are perched on emergences like those of 

 G. pectinata ; becoming woody after the scales have fallen away, they form the 

 well-known "armature." The connecting link is G. pectinata, whose branched 

 bristles may well be the prototype of the chaffy scales (Fig. 475, p. 196). In 

 this character also the Cyatheaceae appear to be the most advanced. 



Turning to the structure of the sorus, the position of which is uniformly 

 superficial, the progression is as in Ferns generally, towards a larger number 

 of sporangia individually smaller, and with a smaller spore-output from each. 

 The Gleicheniaceae, Lopliosoria and Metaxya are all Simplices, with low 

 receptacle and no protection for the sporangia excepting interspersed hairs. 

 It has been shown how an increase in number of the sporangia in G. pectinata 

 leads to a mechanical dead-lock, resolved in Metaxya by enlarging the area 

 of the sorus. But in the C}'atheaceae it has been resolved by elongation of 

 the receptacle, and the introduction of a gradate sequence of sporangia. In 

 Alsophila there is no protection of the young basal sporangia, but in 

 Heniitelia an incomplete basal indusium appears on the side of the sorus 

 next the midrib, while in Cyathca it is completed as a circular cup. This 

 indusium may be held as a new formation, and the biological value of it is 

 obvious in high-growing Tree-P"erns with exposed leaves. The indusium of 

 the Schizaeaceae was also a new formation: there is no reason to assume that 

 such a development can only happen once. It is this assumption that seems 

 to explain the comparison by von Goebel of the two-lipped indusium of the 

 Dicksonieae with that of the Cyatheaceae {Organographie, II. Aufl., p. 1 148, 

 Fig. 1 142). In view of the sequence here submitted, and supported along so 

 man}' lines of detailed comparison of form and structure, it would appear 

 more probable that in two distinct types, the one with superficial sori and 

 profuse scales, the other with marginal sori and hairs but no scales, the 

 protection of the sori was secured by independent origin of indusial growths, 

 more or less homoplastic but not homogenetic. 



In the series under discussion the sporangium itself shows a progressive 



