113 
The first task, therefore, became, through careful investigations of the sexual power 
of reproduction of the forms, to try to determine and to eliminate the hybrid products 
which, until recognized as such, here as elsewhere might supposedly be a main cause of 
the difficulties encountered in the classification. This task entailed, it is true, a large 
amount of work , 1 but at the same time a remunerative one. It became apparent that 
those types, the pure ancestry of which could not be doubted, always w T ere characterized 
by a normal power of reproduction, whereas such forms which, on account of their manner 
of occurrence and morphologically intermediate position, one might be tempted to interpret 
as hybrid products, also were stamped as such biologically, inasmuch as in them, equally 
constantly, the formation of embryo as well as the production of pollen capable of effecting 
fertilization proved either quite abortive or highly reduced. 
After thus the hybrid products, without any real difficulty^ had been eliminated, and 
when furthermore a number of so far overlooked differences m the construction of the 
flower had been found to furnish some exceedingly valuable distinctive characteristics, 
the different types 2 stood out with a clearness that left nothing to be desired.” 
A. tenuis Sibth. ( a distinct species) 
The forms of Euagrostis which, as has been pointed out in the preceding, 
by some North American botanists are considered as belonging to one 
single species and by others as representing three or even more, are easily 
divided into two natural groups, separated as follows: 
(1) The ligule of the lower and middle leaves elongated, 2-5 mm. long, 
rounded at apex; palea shorter than lemma. 
(2) The ligule of the lower and middle leaves very short, 0-5-1 -3 mm. 
long, truncate at apex; palea about | as long as the lemma. 
To the first group belong the forms which by North American botanists 
have variously been called A. stolonifera L., A. alba L., A. palustris Huds., 
A. maritima Lam., and A, alba var. maritima (Lam.) Meyer. 
To the second group belong the forms which have been called A. tenuis 
Sibth., A. vulgaris With., A. capillaris L., and A. alba L. var. vulgaris 
(With.) Thurb. 
As a rule, there is little difficulty in placing a given form of Euagrostis 
with one or the other of these groups, but, on the other hand, specimens 
are not uncommonly encountered which appear to be quite intermediate 
and which form, as it were, connecting links between them. These inter- 
mediate forms, however, occur erratically and generally speaking under 
conditions to make them suspected of being more or less accidental products. 
They invariably display, as Murbeck (1. c.) has shown, and as has been 
corroborated by later students of Agrostis, a very high degree of sterility, 
with at most 10 per cent of the pollen sufficiently well developed to be 
capable of effecting fertilization, provided that the anthers opened, which 
as a rule they do not do, and with no formation of seed at all. The sterility 
of these morphological intermediates clearly points to their origin as a 
result of intercrossing or hybridization of individuals belonging to different 
Linneons. With them eliminated, the A. stolonifera and A. tenuis groups 
are sharply defined and distinctly separated by the characters given in 
the above. The two groups must, therefore, be considered as representing 
different Linneons. In other words, the forms variously called A. tenuis, 
A, vulgaris, A. capillaris, and A. alba var. vulgaris, must be regarded as 
specifically distinct from the ones belonging to the A. stolonifera group. 
1 "In order to arrive at full certainty, pollen examinations were made, aside from on living material, on about 
1,600 herbarium specimens belonging to the National Museum and to the botanical museums of Lund and Upsala 
whose collections kindly were placed at my disposal.” 
2 i.e. species in the Linnean sense, or Linneons according to Lotsy’s terminology. 
56986-8} 
