XII] INDIVIDUALITY OF THE SORUS 225 



Ferns now living. Its relation to a widening leaf-expanse has already been 

 noted. This suggests that a like progression, executed very early in their 

 evolution, may probably account also for the superficial position of the sori 

 in the Marattiaceae and Gleicheniaceae. The facts observed in the living, but 

 very ancient Osmundaceae, combined with the prevalence of the marginal 

 position in most of the early types of Ferns, support the suggestion. Accord- 

 ingly it may be held as probable that a distal or marginal position of the single 

 sporangium was general in the first instance for the Filicales ; but that as the 

 leaf-blade expanded the marginal sorus, zuhether monangial or composed of 

 many sporangia^ slid to the lower surface, and that this has Jiappened along 

 many distinct pJiyletic lines, sometimes early, sometimes late in their Evolution. 



The deeply intramarginal position of many sori of Davallioid Ferns, especially where 

 they are seated on a vein-ending as in Arthropteris and others, suggests a possible explana- 

 tion of the anomalous sori found on the upper surface of Deparia Moorei (see J. M. 

 Thompson, Tra?ts. R. S. Edin. Vol. 1, p. 837, where reference is made to other writers). 

 Normally the sori are marginal, as they are in the allied D. prolifera. But occasional sori 

 are found on the adaxial surface of the leaf, without intermediate positions bridging the 

 gap. It is possible that such sori are initiated in a new and independent position, as in 

 Aspidiian (P.) attomaliim (see p. 217). But the fact that intramarginal sori seated on vein- 

 endings are frequent in the Davallieae,to which {:^ivc\\\y Deparia should be assigned, suggests 

 that its anomalous sori may be referred to a like origin with them. 



Loss OF Individuality of the Sorus 

 The Sorus, made up as it usually is of a compact group of constituent 

 parts, viz. the receptacle, sporangia, and indusial flaps, appears in so many 

 Ferns to be strictly circumscribed that the natural tendency is to regard it 

 and to describe it as an entity : and this view is accentuated by the recog- 

 nition of the changes in its position just described. For isolated examples 

 of Ferns it may be legitimate to hold it as an entity. It is stamped as 

 such by nutrition. Its limited outline appears to be dictated by the limit of 

 convenient radius of nutritive supply from a vein-ending, marginal or super- 

 ficial, on which it is seated. But if this be the real cause, then it may be 

 expected that a bifurcation of a vein will produce tWin sori : or that a fusion 

 of veins, as seen in so many of the later reticulate types of Ferns, might be 

 expected to result in fusion-sori: or these again may break up into fragments 

 in case the vascular arches were again interrupted. It will be seen that all 

 of these states can be illustrated in living Ferns : and other facts will be 

 adduced which show that the sorus is not a morphological entity for Ferns 

 at large. 



A primitive state for Ferns was no doubt that in which the leaf or segment 

 was narrow. Here it is commonly found that a single row of sori, marginal or 

 superficial, is disposed on either side of the midrib, and this arrangement of 

 them appears to be fundamental. It is seen in such primitive types as the 



B. IS 



