ZOOLOGY AND BOTANY, MICROSCOPY, ETC. 199 



have obviously been differentiated from a common parent-type by pre- 

 vailing environmental conditions in the several life-areas in question, 

 such as is so often seen among vertebrates. At the same time, many 

 closely related variants do exist among plants, differentiated to the same 

 varying extent as in the geographic races of birds and mammals, but all 

 occurring in the same life-zone, and often side by side. They are, more- 

 over, quite constant in their racial characters, and are not cases of 

 individual variation. The agency responsible for this differentiation is 

 in many cases to be found in the varying soil conditions and other local 

 peculiarities not strong enough to affect higher animal life. Owing to 

 the fixed nature of plant life, such conditions may have a very marked 

 effect in producing local forms ; to these the author would apply the 

 trinominal nomenclature, to distinguish them from the clearly defined 

 species. The author points out at the same time that nomenclature 

 becomes absurd when applied to variants which can only be recognised 

 by one or two specialists, who have devoted years to the study of the 

 group, as has happened in the genus Gratagus. 



The species of Viola fall into two groups, the caulescent and the 

 acaulescent ; those of the former show but little of the tendency to 

 racial variation which characterises the latter group. Three types of 

 colour — blue, yellow and white — occur in each group; parti-coloured forms 

 also occur in the caulescent. The blue-flowered acaulescent species 

 show by far the greatest racial and individual variation, and it is upon 

 these that the author bases his general statements. Leaf -form supplies 

 the most striking character. There is often a marked similarity in the 

 general shape of the early leaf in a number of species, which is more 

 or less entire, passing later into a lobing characteristic of the species. 

 Starting from the most primitive type, the cordate leaf, the author 

 indicates various racial variations, tending in the one direction towards 

 extreme lobing or leaf-division, in the other towards the triangular and 

 narrow sagittate-leaved forms. There seems to be a tendency towards 

 narrow leaves in many wet-ground species. Variation also occurs in the 

 degree of pubescence of the leaf. In using the relative length of petiole 

 and scape comparison should explicity be made with either the first or 

 second set of leaves, as the flowering period often covers the growth 

 of the second leaves, so that early flowers are longer than the leaves, 

 while later ones are shorter. The length and character of the peduncle 

 of the later cleistogamic flowers is an important specific character ; 

 except that in all wet-meadow species it is erect, no correlation is 

 possible between this character and the nature of the habitat. In floral 

 characters there is a great amount of individual variation. The extent 

 of pubescence on the petals is an important specific character. 



As regards geographic distribution of the forms found, both acaule- 

 scent and caulescent, in eastern Pennsylvania and southern New Jersey, 

 six are characteristic boreal species ; three others also occur in the 

 higher Alleghanies. Another group seems to be decidedly southern. 

 The other forms range indiscriminately over the intervening country. 



The author gives a clavis of the species (thirty in number) and 

 forms, based on the above-mentioned characters ; this is followed by 

 descriptions, including the range and habitat of the individual species 

 and forms. 



