A survey of the Phytogeography of the Arctic American Archipelago 37 
Trevelyän, and wrongly stated to come from Melville Island. They are, indeed, 
recorded in the foUowing, but I have pointed out which I look upon as doubtful 
and excluded them frora the phytogeographical discussion at the end. 
To go into particulars for the synonymics of each species I have not thought 
necessary. For the largest part of the flora this is to be found in my work, quoted 
above, or in my other paper Plants NW Greenl. In a few cases, indeed, it has seemed 
advisable to add some notes about synonymics in general, in inost instances only 
the names are mentioned under which the plant appears in works about our area, 
if they are different from those used here. As for the circumscription of the spe- 
cies, I am fully aware that it will be critized by some readers as far too wide. 
But I have reasons for keeping up a conce])tion of species, nearly coinciding with 
the original Linnean one, and as these seem to me rather weigthy, I shall state 
them here. The modern tendency of establishing «small species* certainly has in 
many cases led to valuable results in deepening our views upon the origin of new 
forms, and from a phytogeographical point of view the splitting up of old widely 
circumscribed species has led to a better understanding of the degree of relationship 
between different floras and thrown light upon the questions of immigration, but 
on the other hand it also has contributed to make phytogeograi:)hical researches 
more difficult and to obscure features which ought to be easier recognizable. 
The « small speciess estaWished in låter times may, I think, be separated into 
three different groups. In the first of these I place the new species discerned within 
the range of an old species, a group of species, or a genus, after a close and con- 
scientious study of the plants in questiou in their natural habitat as well as in 
herbaria. Thus hereditarily constant forms, distinguished by a number of charac- 
ters which are combined without excei)tion are separated as species. As giving 
excellent examples of «small species« of this kind the Alchemilla-stuåies of Buser, 
MuRBECK, and Lindberg may be looked upon. In such cases the different constant 
types may appear together in the same station or at least their areas may coincide 
more or less. The second group of « small species* is that of the geographical 
types, viz., of forms well circumscribed and breeding true within a certain area, but 
raerging gradually into others where their geographical range is not surrounded by 
more or less impassable borders, such as wider expanses of sea or high mountain 
chains. A great many good samples of this kind of species we find in the Euphra- 
.virt-raonograph of Wettstein. When, as in the case mentioned, the estabhshment 
of species is based on comprehensive and conscientious studies, the result will be 
very valuable and apt to throw light on phytogeographical questions. But in less 
conscientious hands this method may easily result in the establishing as species of 
forms which would have been far better placed in a lower syste.natical rank. 
The third' group of «small species» is formed of all such as are singly or in 
greater or lesser numbers created on an incomplete material, for instance on a small 
collection from a region of which the flora and the variation of each separate spe- 
cies is more or less unknown to the author. Even in such a case he may, of course 
