118 Rhodora [JULY 
paper. In addition, the writer has seen the material in the Herbarium 
of the Missouri Botanical Garden, the herbarium of the Philadelphia 
Academy of Natural Science, the National Herbarium, the herbarium 
of the New York Botanical Garden, and the herbarium of Cornell 
University. The principal cause of most of the difficulty experienced 
by students of Amelanchier in the past in drawing specific lines, is to 
be found undoubtedly in the wonderful variability which seems to 
exist within the genus, a variability so extreme that rarely did it seem 
possible to find two plants as much alike as specific identity in other 
groups would demand. Indeed, if the sheets of any large collection 
are sorted on any assumed basis, almost as many transitional sheets 
will be found as sheets typical of the supposed species. For this 
reason many taxonomists have been inclined to believe that no real 
specific lines exist, at least so far as the eastern species are concerned. 
Because of this wealth of form two courses have seemed possible; a 
conservative treatment in which a very few species, sometimes only 
one, may be recognized; or a radical one in which a great many species 
are supposed to exist. 
In attempting any treatment of the genus the first point to be de- 
cided is the significance of this wealth of form. In the past it has 
been usually interpreted as real variation, and, if such, we must admit 
that to draw specific lines would be impossible. Recent activity 
in the study of heredity has shown, however, that hybridization in 
nature, especially in certain groups, is much more common than was 
supposed. Moreover, the Mendelian theory has supplied a method 
of testing supposed hybrids by growing the offspring through several 
generations. By means of such tests it has been shown in many cases 
that so-called intermediates are due to hybridization. It has also 
been shown that many so-called intermediates are really notinter- 
mediates as far as the individual characters are concerned, but are 
made up of the characters of the two parents combined according to 
the law of combinations and permutations until all possible combina- 
tions exist. The question may be legitimately raised, therefore, as to 
whether the multitudinous forms in Amelanchier are hybrids, and, if 
so, as to what are the true species. 
The writer began the present study under the belief that the wealth 
of form was due to true variation; but with more material at hand, 
and especially in recent years since the possibility of hybridization 
has been more in mind, a surprisingly great number of cases has been 
P. 
