112 
Species and Hybrids 
The term “species” is in many cases applied more or less at random 
and, as a consequence, in very many cases to systematic units of mani- 
festly widely different rank. In this connexion, however, it is of less 
importance to argue what the term “species” should or should not imply 
than to set forth how it is applied, i.e., to make clear what systematic units 
the writer has in mind when speaking of “species”. 
From many years of experience in plant breeding, when thousands of 
forms of wild and cultivated grasses were observed and studied, the writer 
is convinced that, although the plant breeder obviously must operate with 
forms separated by minute characters, it is, for purposes of descriptive 
taxonomic botany, well-nigh necessary to work with comparatively large 
basic units. In the following the term “species” will, therefore, be used 
in a rather wide sense. It will be employed to designate groups of forms 
which, although in several respects differing from each other rather 
considerably, yet have morphological characters in common which clearly 
indicate that they are of a very close systematic relationship. In other 
words, the term species, as here applied, is identical with Lotsy’s “Linneon”, 
a term proposed “To replace the term species in the Linnean sense, and to 
designate a group of individuals which resemble one another more than they 
do any other individuals” (Lotsy 23, p. 27). It is, of course, obvious that all 
forms thus included within the boundaries of a “species” or “Linneon” 
must have some common constant characters separating them from other 
“species” or “Linneons”. 
Where one “Linneon” ends and another, closely related one, begins, 
is often told by Nature herself. A criterion of “Linneons” is that they do 
not usually intercross and that, when they do, their progeny is as a rule 
characterized by a high degree of sterility in both the male and female 
organs. Such products of inter-Linneon crosses, generally termed hybrids 
in descriptive, systematic botany, are not uncommonly met with in grasses, 
although, so far, slight attention appears to have been paid to them by 
North American botanists. Ascherson and Graebner (1) recognize over 
forty grass hybrids, Neuman (26) has thirty-three, Lindman (18) twenty- 
five, and in the last edition of “Catalogue of Scandinavian Plants,” pub- 
lished by the Botanical Society of Lund, Sweden, not less than forty-four 
grass hybrids are listed. In the genus Agroslis several hybrids are well 
knowrn, and in all cases their hybrid nature manifests itself by a very high 
degree of sterility. Thus Holmberg (14, pp. 143-45) mentions four, viz.: 
A. stolonifera x tenuis, A. canina x stolonifera, A. canina x tenuis, 
and A. borealis x stolonifera, none of which has more than 10 per 
cent of perfectly good pollen. These four hybrids w r ere first segregated 
and described by Murbeck (25) who fully realized the importance of 
recognizing them as accidental products of a taxonomic value quite different 
from that of the true species. As Murbeck’s paper, which is written in 
Swedish, apparently has so far been overlooked by North American 
students of the genus Agrostis, the w r riter translates from it, as follows: 
“When, in the summer of 1895, I desired to attempt a more satisfying critical analysis 
of the North European forms of Agrostis it soon became apparent that such an analysis 
would not be possible if an exclusively morphological method of investigation were employed 
and that the reason for this was that production of hybrids played a not unimportant 
role within the genus 
