224 
AMERICAN JOURNAL OF BOTANY 
[Vol. io, 
shepherd’s purse {Bursa bursa-pastoris) and to a less extent with several 
other species of the same genus. In connection with this work I have had 
material sent to me from nearly every part of the temperate regions of the 
world. Every new lot of material received has added to the demonstration 
that there occurs in nature an amazing number of hereditarily distinct 
forms within the species B. bursa-pastoris. While these biotypes are 
genetical entities of perfect delimitation, they can not by any method be 
given taxonomic validity as species or subspecies,- because, under natural 
conditions, the genotypic factors prove to be less effective in determining 
the visible qualities which are necessarily utilized in classification by inspec¬ 
tion than are the environmental factors. Even the larger features of 
leaf-lobing to which I have assigned the subspecific names heteris, rhomboidea, 
tenuis, and simplex are, oftener than not, quite indistinguishable in the 
field, and it has not been infrequent to have forms sent to me by good 
taxonomists, under the name u integrifolia,' 1 ' turn out to be heteris, the most 
highly lobed type. The significance of this is that the distinctive genetical 
characters are easily suppressed by crowding, poor soil, poor illumination, 
and other adverse conditions. The only practicable thing to do in a case 
of this kind is to maintain the species bursa-pastoris, recognizing its com¬ 
posite character, and in so doing to recognize also that species must be 
based upon characters whose existence is not too seriously affected by 
deviations in the environment. 
In relation to the question of sterility of hybrids as evidence of specific 
distinctness in the parents, I may cite the following cases in Bursa bursa- 
pastoris, and its hybrids with B. grandiflora and B. Viguieri. 
In 1913 I received from Professor A.H. Trow, of Cardiff, Wales, several 
packets of seed of Bursa specimens growing in close proximity to one 
another at Cardiff. Two of the progenies grown from these seeds differed 
slightly in outline of the capsules and in form and surface of the leaves, 
but neither suggested by its appearance that it ought to serve as the type 
of a newly defined species, and, indeed, it is doubtful whether they could 
have been separated at all when growing in nature. However, the two 
strains thus derived, when crossed together, gave hybrids which were 
almost completely sterile. When two types are thus found intermingled 
in nature and not capable of ready separation on inspection, they can not 
properly be referred to distinct species, even though their crosses show 
sterility; but, on the other hand, when we find that all up and down the 
Pacific coast of North and South America and extending eastward at least 
as far as Tucson, Arizona, and Waco, Texas, a characteristic form of Bursa 
occurs, having invariably concave-sided capsules, and yielding only sterile 
or almost sterile hybrids when crossed with numerous biotypes of B. bursa- 
pastoris from Europe and the eastern half of North America, which latter 
biotypes have prevailingly convex-sided capsules, one may with considerable 
justification urge the recognition of this west-American form as a species, 
