38 HARPER 



relation to environment from non-evergreens, annuals from per- 

 ennials, epiphytes from parasites, etc. By the classification here 

 adopted the dominant members of any habitat-group are at once 

 distinguished from those which are more or less dependent, such 

 as vines, epiphytes, and parasites. By making evergreens con- 

 spicuous as I have done the difference between winter and summer 

 aspects of each group is shown at a glance. Some striking dif- 

 ferences between different habitat-groups are also brought out 

 in this way. 



It is interesting to note that all epiphytes and bryophytes (in 

 our territory at least) are evergreen. 



Second : keeping the four main classes distinct (for it would not 

 be practicable or fair to compare trees with shrubs, shrubs with 

 herbs, etc.,) the species in each class are arranged as nearly as 

 possible in order of abundance. In all but some of the rarest hab- 

 itat-groups I have placed before each name a number correspond- 

 ing to the number of times I have definitely noted that species 

 in that particular habitat. This gives approximately the rela- 

 tive frequence, which in most cases is very nearly the same 

 as the relative abundance. Some phytogeographers in recent 

 years have undertaken to determine relative abundance by actu- 

 ally counting the individual plants on small measured areas of 

 ground. While this method leaves nothing to be desired as far 

 as accuracy is concerned, it would take an incalculable amount 

 of time to apply it to any considerable portion of the region under 

 discussion, and then the final results might not differ much from 

 those obtained by the simple method adopted here. In the case 

 of a few very abundant species I have not taken the trouble to 

 note them in the field as often as some rarer ones, but it is easy 

 enough to place these in their proper places at the head of their 

 respective lists. 



Furthermore, after each species, if it is a flowering plant, I have 

 indicated its normal flowering period, as far as known, by figures 

 representing the months. (In the case of the vascular cryptogams 

 these figures are replaced by o.) And if the flowers are ento- 

 mophilous the predominating color of the corolla (or other organ 

 which serves to attract insects), is indicated. For this purpose 



