Descendenz und Hybriden. 339 



Following a short analysis of racial Variation, — in which 

 it is stated that the author's studies do not Warrant an attempt 

 to account ior the origin of local plant races, although they are 

 just as cleariy differentiated as the geographic subspecies of 

 vertebrates, and should be designated by trinomials, — a par- 

 ticular account of Variation in the genus Viola is given, and 

 this is followed by an analysis, with füll descriptions and notes, 

 referring to the species of Viola of Philadelphia and its 

 vicinity. Of these, 30 forms are recognized. A bibliographic 

 index to the species is appended, and the plates represent leaf 

 outline and growth habit. 



The following new names are included in the paper: V. 

 cuciillata macrotis ( V. macrotis Greene) , V. ciicnllata lepto- 

 sepala (V. leptosepala Greene), V. palmata variabilis (V.varia- 

 hilis Greene), V. palmata angellae (V. angella Pollard) and 

 V. fimbriatula aberrans (V. aberrans Greene). Trelease. 



Tammes, Miss T., On the influence ofnutritiononthe 

 fluctuating variability of some plant s. (Proceed. 

 Kon. Akad. v. Wetensch. Amsterdam. Dec. 24, 1904. 14 pp. 

 1 PI.) 



In the Summer of 1903 some experiments were made in 

 the Botanical Garden in Groningen with 7 plants, of which 15 

 veral characters were examined statistically, in widely different 

 conditions of nutrition. 



The principal results are the following. Nutrition influences 

 the median value and the variability of the characters. The 

 median value augments with good nutrition and its sensibility 

 has been calculated by dividing the difference of the median in 

 the well-fed and the badly-fed plants by the value for the well- 

 ied plants. The value thus obtained, and here used for the 

 first time, is called sensibility-coefficient. This coeffi- 

 cient is very different for different species and for different 

 characters of the same species. 



The variability-coefficient q/M is with good nutrition fairly 

 constant for different characters of the same species, but very 

 divergent for the different species. With bad nutrition two of 

 the species studied showed great differences between the varia- 

 bility-coefficients of the characters of the same species. The 

 sensibility-coefficient of q/M diverges greatly for different species 

 and characters; for some characters it is positive, which means 

 that good nutrition results in an increase of variability^ for other 

 characters, even of the same species, it is negative. 



For two characters the symmetrical curve on fertile earth 

 was altered by bad nutrition to a semi-curve, and in one of 

 these cases the origin of this phenomenon could be explained. 



Moll. 



22* 



