24 



For it is well known that genera are more prolific of new forms 

 at the edge than at the center of their range. A striking case 

 in our area is that of Panicum and Carex, which both have dozens 

 of specific outposts in the region. As it happens none of these 

 are endemic here, but that in no way invalidates the contention, 

 as the species may spread as in both these cases, so as not to be 

 endemic in our particular region. 



In somewhat the same category are those endemics which, 

 while they do not belong to genera that dwindle in our region, 

 are related to species that find their distribution outposts here 

 or very near here. Such are Dentaria incisifolia (related to D. 

 maxima) , Hibiscus oculiroseus (related to H. Moscheutos) , Stachys 

 atlantica (related to S. hyssopifolia) , and Pyrola oxypetala (related 

 to P. chlorantha). 



Here again, as in the endemics that are found in the dwindling 

 genera, it is the element of instability, which seems to be the 

 underlying cause of the production of new forms. There may 

 have been thousands of such forms produced in our area but 

 only a few have been detected. Some of these are already wide- 

 spread, but it does not seem clear that they are therefore, as 

 Willis would have us believe, the most ancient of the endemic 

 element. For the speed with which the endemic spreads may 

 have little or nothing to do with the cause of its origin. So 

 many factors enter into the distribution of an endemic, that it 

 seems scarcely safe to take this as a criterion of its age in any 

 given area. Other things being equal, we should expect to 

 find our newest endemics the most restricted, and vice versa. 

 But other things are not equal, indeed the inequality of other 

 things, is, as we have seen, so great that such a criterion of 

 antiquity in endemism must be used with extreme caution. 



In the whole native flora of the region about 20 per cent, of 

 the species reach their distribution outposts within the area. 

 Excluding the relict endemics, noted above, all the rest of our 

 endemics except three to be treated presently, belong to genera 

 that dwindle, or are related to species that are on or near their 

 outposts, in this region. The inference that generic and specific 

 instability is the chief factor in the origination of these endemics 



