64 BASIS FOR SYSTEMATIC TREATMENT OF FERNS [CH. 



instances. The scope of variation in respect of it must be examined, and 

 extreme types recognised. It will then be a question in relation to each 

 criterion how far the extreme types can be held as relatively primitive or 

 as advanced. Such questions in the case of each criterion can be resolved 

 with some degree of probability by comparison in respect of other characters. 

 But more material assistance to a conclusion is afforded by reference to the 

 geological record. For instance, some living Ferns are grouped as Euspo- 

 rangiate, that is, they have relatively bulky sporangia: others as Lepto- 

 sporangiate, being characterised by relatively attenuated sporangia. It is 

 now known that the Ferns of the Palaeozoic Period had relatively bulky 

 sporangia, corresponding to the Eusporangiate Ferns of the present day. 

 It is even a question whether true Leptosporangiate Ferns existed in the 

 Primary Rocks, though the)^ constitute the bulk of the present-day species. 

 The natural conclusion is that the Eusporangiate type is relatively primitive 

 and the Leptosporangiate derivative. Again, an open venation, without fusion 

 of veins, is characteristic of the earliest fossil Ferns. Fusion of veins is first 

 seen in the Middle Coal Period, and a small-meshed reticulum appears only in 

 Ferns of the Mesozoic Era. The natural conclusion is that open venation is a 

 primitive, and a small-meshed reticulum a derivative character. Such con- 

 clusions commonly running parallel to and supporting one another, produce 

 a cumulative effect. This may even help towards decisions respecting yet 

 other criteria. And thus by coordinated study of the whole series of criteria 

 the web of evidence may be ever more closely drawn. The result in favour- 

 able cases will often be a confident opinion as to the seriation of distinct 

 but related families, genera, and even species according to descent. In less 

 favourable cases a reasonable probability can usually be established. 



As a further test or corroboration of the conclusions thus acquired the 

 method of observation of the successive steps in the individual life may be 

 used. For instance, the outline and venation of the earliest leaves of the 

 sporophyte are liable to differ from those formed later by the same plant, 

 showing characters more primitive than those of the adult leaf It is, how- 

 ever, in the study of the vascular tissues that this line of ontogenetic evidence 

 finds its best opportunity, though owing to singular neglect in collecting the 

 necessary facts by serial sections it has not hitherto been developed as it 

 should have been. The ontogenetic development is in point of fact a natural 

 key by which to interpret the stelar elaborations so characteristic of the 

 Filicales. But it can only be applied subject to limitations, since the ontogeny 

 is often abbreviated, and steps that might have been anticipated are frequently 

 condensed, or even omitted from the individual life. 



Following the lines thus briefly sketched the first duty will be to examine 

 each of the criteria of comparison critically. A Chapter or more will be 

 devoted to each. The nature of the variations within the Filicales will be 



