XLiii] COMPARISON i6i 



vary, as it does in Cyathea sinuata (Vol. II, Fig. 557), and in M. stnithiopteris 

 (Fig. 678, D, K): on the other hand the sori may be in such positions as to 

 form regular intra-marginal series. This is common in the Cyatheaceae, and 

 particularly in species o{ Alsophila and Hemitelia. It is found in Matteuccia 

 orieiitalis and intermedia (Fig. 680), and it becomes a constant and indeed 

 an essential feature in Loniaria. 



The distinctively gradate condition of both Matteuccia and Onoclea points 

 directly towards the Cyatheaceous type. The detail of the large sporangium 

 of M. intermedia, more complex than that of Onoclea as represented by 

 Campbell [I.e. PI. Vli, Fig. 26), compares most nearly with that o{ Lophosoria, 

 those of the other species being smaller. It is significant that the larger and 

 more complex sporangium goes along with the absence of an indusium in 

 M. intermedia, though this protection is present in the rest. These features 

 indicate M. intermedia as the most primitive member of the family, while 

 they direct comparison towards LopJwsoria and Alsophila among the 

 Cyatheoid Ferns, The inconstancy of occurrence of an indusium within a 

 very close and natural circle of affinity raises again the question of its 

 nature and origin. We have seen in Chapter XL how variable is the occur- 

 rence as well as the form of the indusium in the Cyatheaceae, and how 

 Christ found, even within the genera Alsophila and Cyathea, that the absence 

 or presence of an indusium is not a generic constant. The same now applies 

 in Matteuccia, which includes both indusiate and non-indusiate species. In 

 the case of M. intermedia there is no evidence of abortion to explain its 

 absence: the structure of the young sorus, as seen in Fig. 6'^'^, is as consistent 

 with an originally non-indusiate state as are those oi Alsophila (Vol. il. Fig. 

 564), or Lophosoria {Ann. of Bot. XXVI, PI. XXXV), or even of Gleichenia itself 

 (Fig. 490).' In none of these is there any reason for entertaining the idea of 

 abortion to explain the non-indusiate state. The structural evidence indicates 

 rather the advent of the indusium as a new structure. It may be held to have 

 made its appearance de novo in M. stnithiopteris and in Onoclea : but 

 M. intermedia shows by its absence a primitive condition, which it shares 

 with the series of Superficial Ferns above named. The existence of this 

 second instance of inconstancy of an indusium in relatively primitive Ferns, 

 related to but distinct from the Cyatheaceae, strengthens the conclusion that 

 an indusium can be formed de tiovo, while conversely it discounts any forced 

 comparison with the lower indusium of the marginal series as represented 

 in the Dicksonioid Ferns. 



The next Chapter will deal comparatively with the Lomarioid Ferns. It 

 will there be apparent how important a part in the probable evolutionary 

 sequence is taken by the Onocleoid Fern Matteuccia intermedia: for the view 

 will be advanced that the Lomarioid Ferns were derived from some such 

 source as that represented by this new non-indusiate species. Meanwhile the 



