PROCEEDINGS OF SOCIETIES. 143 



in various localities. There are some characters, the uniformity of which, of 

 primary importance in considering genera and species, are but secondary in con- 

 sidering variety — such as character of venation, division of frond, fertility ; these 

 are variously modified, according to the type of variety or subvariety, and, of 

 course, to a certain extent, modify it in return. More generally we find the variety 

 barren than the subvariety, but this may arise from some mistake of manipu- 

 lation in cultivation. Indeed, for my own part, I believe most varieties may be 

 made to produce fertile fronds, at least in the two forms under consideration to- 

 night. 



All the groups thus formed are capable of general definitions, though it must not 

 be expected that this general definition will detail every minute feature of variety 

 found in the many analogous forms necessarily included under it, any more than 

 the definition of species describes every minute feature of the individual plants com- 

 prised, these being of little importance, provided the general characters are adhered 

 to. To each of these groups a name is proposed to be given, by which it may be 

 known, so that, for most species, it will suffice to quote this name to show what the 

 form of the variety is. When two or more distinct forms of the same variety occur 

 under a species, it will be but necessary to mark them in some way, as with Greek 

 letters, &c, after each, and describe the salient points of difference. Ex. gr., under 

 Ath Filix faemina, we find three forms of a variety which we will call ramosum ; 

 these could be thus described — Ath. Filix faemina, var. ramosum (three forms). 



A, rachis branched ; pinnules decurrent ; apices of segments curled. B, rachis 

 split into numerous linear segments; pinnae contracted ; apices of pinnae split into 

 linear segments, tasselled. G, rachis and pinnae multifidly crisped at apices, and 

 tasselled. This is the only species in which the forms of this variety cannot be re- 

 ferred to one description, and when the species of Athyrium are definitely arranged, 

 possibly the forms will be found to belong to different species. ( Vide second part of 

 paper.) To enter into all the subvarieties and varieties would detain you too long. I 

 shall, therefore, content myself with one of each, taking those which are the commonest, 

 and which, fortunately, also happen to be the most fully worked out. These are — 

 var. ramosum and subvar. multifidum. These both consist in a repetition of the various 

 parts of the normal frond — this is a true duplicature, not a mere modification of other 

 organs such as we see in double flowers. Every part of the aerial organs is found 

 doubled ; some more commonly than others. In simple fronds we find it in the 

 stipe, the rachis, and the apex of the frond. In compound fronds, in addition, it is 

 also found in the pinnae and the pinnules — in fact, it may be found in every axis of 

 the plant, primary, secondary, tertiary, &c. It is also found in the sori, though 

 more difficult to demonstrate, as in some forms of Phyllitis scolopendrium ; I met, 

 myself, with a wild specimen of Trichomanes speciosum, which I now show ; when 

 I got it almost all the thecae were in pairs. 



Perhaps it is wrong to call the organs doubled, as they are not strictly so all 

 through, but arise single, and then split into two, obeying a law of nearly universal 

 occurence in nature, in accordance with which we find that the aberrant forms of 

 the higher groups copy closely the normal type of those below them ; thus, the 

 dichotomy, which obtains in these forms, is the normal state of some, at least, of 

 the Lycopodiaceae — ex. gr., Selago, and is also seen in the seed lobes of some 

 exogens. 



Some have proposed to separate into different groups the forms placed at the 

 head of these groups — i.e., where the division affects the stipe, but this appears con- 

 trary to analogy. They ground their opinions chiefly on its rarity of occurrence 

 among species, but even they allow that it is not a constant character ; we will find 

 it is rare, but not so rare as either the division of pinnule, or sori, or thecae ; and 

 if we do not make any distinction between the rachis and stipe — which division, 

 after all, is an arbitrary one — we will find the form not so rare at all — in fact, the 

 nearer we go to the apex of the primary axis, the commoner will we find this 

 division ; and it is what might be expected, the modifying causes not having an 

 opportunity to develop their action till the frond is partially developed, for if it acted 

 previously, we would have two fronds instead of one. If we examine a simple frond 

 we will find this well shown, as the rarest specimens are those divided deeply, and 

 the commonest those having merely the apex split, often multifidly so. The 



