117 
the shore of the sound between Sweden and Denmark. In Canada, the 
number of forms which so far have been found is comparatively small, a 
fact which is quite natural when it is considered that most of the forms of 
the stolonifera group, if not all, are introductions from Europe. Yet, in 
some districts, for instance on Vancouver island and in the Maritime 
Provinces, particularly Prince Edward Island, a not inconsiderable number 
of more or less distinct forms are growing. 
In order to determine whether the various forms are to be regarded as 
species, with the definition given in the preceding, or as varieties of one 
polymorphous species, it is necessary to ascertain to what extent the 
characters separating them are to be considered “essential.” 
The main characters involved refer to the shape of the panicle, the 
presence or absence of stolons, and the presence or absence of awn on the 
lemma. 
The panicle varies from narrowly contracted and spike-like (Plate 
XVI, figure 2; Plate XV) to ample and spreading (Plate XIV, figure 1; 
Plate XVII). Between these two extremes all kinds of intermediate 
panicle types are found. It should be explained that the types in question 
refer to shape of the panicle as it remains after flowering, and not to its 
appearance during flowering time. 
The question is now: can the various panicle types be considered of 
sufficient taxonomic value to be used as species characters alone? In 
order to answer this question, it is advisable, if not necessary, to examine 
the question from the standpoint of panicle variation in grasses in general. 
As a first exhibit in the argument, as it were, Dactylis glomerata L. 
may be used. This species, as Witte (44, figs. 25-30) has shown, is com- 
posed of an exceedingly large number of forms, differing from each other in 
respect to practically all vegetative characters. Witte figures forms with 
very narrow, almost spike-like panicles, forms with ascending panicle 
branches, forms having the main branches horizontally spreading or 
reflexed, and forms with slender, drooping branches. This tremendous 
variation of Dactijlis glomerata is so well known that it seems unnecessary 
to enter into further details. It may be mentioned, however, that Dactylis 
glomerata, an Old World species and consequently occurring in the largest 
number of forms there, is also found highly variable in suitable localities 
on this continent. Such a locality is, for instance, Victoria, B.C., where 
the writer has observed hundreds of different forms, having a range of 
variation, as to the construction of the panicle, even greater than that 
described by Witte. 
As a second example may be mentioned Bromus inermis Leyss. This 
species, also an introduction from Europe, which has been studied par- 
ticularly in respect to variability at the Experimental Station at Fargo, 
North Dakota, is, like Dactylis glomerata, exceedingly polymorphous. 
The panicle varies from narrow and contracted to ample with widely 
spreading branches, the various forms belonging, as it were, to two main 
series, one with the panicle branches one-sided (Plate XVIII, bottom 
row), and the other with the branches all-sicledly arranged (Plate XIX, 
bottom row). The forms illustrated in Plates XVIII and XIX were all 
collected by the writer in 1922 in a small waste place in the city of Ottawa, 
Ont. Similar variation in the shape of the panicle has been observed by 
the writer in colonies of Bromus inermis elsewhere, e.g. at Saskatoon and 
Regina in the province of Saskatchewan. 
