i9oi- CoLGAN. — Notes on Irish Topographical Botany. 239 



Those who are interested in questions of distribution can 

 readily apply this method to the various other botanical 

 divisions of Ireland, bearing in mind, however, that trust- 

 worthy results can only be looked for in the case of areas 

 which have been fairly well explored. Unfortunately, the 

 materials are still wanting for the working out of such 

 indices for Great Britain, where the variations in the con- 

 ditions are often much wider than in Ireland. The weakest 

 point in this statistical method is obviously the uncertainty of 

 the data as regards introduced species. There are many 

 instances in which perhaps no two botanists would agree on 

 the question whether a given alien plant has so far established 

 itself in a given area as to deserve enumeration in its 

 permanent flora. Where, however, the data for the various 

 areas compared are selected by one authority, as is the case in 

 Irish Topog7 r aphical Botany, this uncertainty is of less moment, 

 as we may assume that the severity or laxity of judgment has 

 been applied to all divisions with a fair approach to uni- 

 formity. 



It may be asked how the results just deduced from an 

 examination of Mr. Praeger's book would compare with those 

 to be drawn from the second edition of Cybele Hibemica. 

 A little reflection will make it clear that no very close 

 correspondence between the diversity index for any pair of 

 the Cybele Districts and for any pair of the included comital 

 divisions need be expected. Floral diversity depends largely 

 on diversity of physical features in the areas compared. If 

 one area be flat and the other mountainous, if one be rich and 

 the other poor in lakes and rivers, if one have large tracts of 

 limestone and the other have none but primitive rocks, these 

 diversities will inevitably find expression in the floras. But 

 since physical differences tend more and more to disappear as 

 the areas increase, the index of diversity for two Cybele 

 Districts should, as a rule, be less than that for a pair of the 

 included counties. And for the same reason, wherever 

 climatic influence is preponderant, that influence should more 

 clearly express itself in the index of Districts than in the 

 index of included comital divisions, 



An instance or two will serve to illustrate this. Taking as 

 data for the Districts the Distribution Table of the second 



