6o BASIS FOR SYSTEMATIC TREATMENT OF FERNS [ch. 



employed to approach that end may be none the less sound in itself, and the 

 attempt may find its justification in a partial success. 



A knowledge of the whole life-cycle, such as is given in Chapter I, is 

 necessary in order to select those criteria which are to be specially used in 

 comparison. Since that cycle is essentially similar for them all, the Male 

 Shield Fern {Dryoptevis Filix-inas, Rich.) will serve as a fair example. 

 Moreover comparison indicates that it is phyletically neither one of the most 

 primitive nor one of the most advanced of Ferns, and is on that account all 

 the more suitable. Its life as detailed in Chapter I will therefore give a mean 

 example of those several features which will be taken as criteria for com- 

 parison of Ferns at large. They will be severally named in the following 

 paragraphs. 



I. The shoot of Ferns though constructed on the general plan seen in 

 Dryopteris is liable to great differences of proportion and of pose. The 

 stem may be erect, ascending, or prone. The number of the leaves and the 

 distance between the individual leaves may vary. The shoot itself may 

 branch in various ways, either at the tip, or by the formation of buds in 

 divers positions remote from the tip. These and other features show that 

 for the Ferns at large tJie form and proportions of the shoot will provide 

 material suitable for comparative treatment. 



II. An examination of the apical cone in Ferns shows that in most cases 

 it is occupied by a single cell of definite form, with other cells regularly 

 arranged around it in such a way as to show that they are really segments 

 which have been successively cut off from that cell. It is accordingly styled 

 an initial cell {¥\g. lo). A similar examination of the tip of the leaf and of 

 the root gives a like result, while somewhat similar cleavages are found to 

 occur with like regularity in the formation of sporangia (Figs, ii, 12). Such 

 segmentations from initial cells are general for all embryonic regions in Ferns : 

 but there are differences of detail in the segmentations, and even in the number 

 and the form of the initial cells. These differences again provide material for 

 comparison, and it will be found that the constitution of the plant-body as 

 indicated by its initial segmentation will serve as a further basis for comparative 

 treatment. 



III. The contours, texture, and venation of the expanded leaves in Ferns 

 show great differences in detail. Usually, as in Dryopteris (Fig. 2), there 

 is a central rachis which bears rows of pinnae right and left, diminishing in 

 size towards the apex. These again are cut into pinnules or secondary 

 branches, while their marginal toothing suggests an imperfect branching of a 

 third order. But on the other hand in some Ferns, as in the Adder's Tongue 

 and Hart's Tongue, the leaves are entire, not being cut into pinnae. Others 

 again may show branching of a higher order than in Dryopteris. There 



