BLASTOGENIC VARIATIONS. 159 



3, 4, 5, and 6 characters were on an average respectively 

 as 10 : 19 : 43 : 78, or very nearly as 1: 2: 4: 8. He 

 also confirmed his law bv some observations on Phase- 



u 



olus vulgaris and P. nanus, but the crossings of P. 

 nanus 9 with P. multiflorus gave only a partial result, 

 whilst those on Hieracium did not agree at all. In the 

 light of the hybridisation experiments of Kolreuter, 

 Giirtner, and others, Mendel recognised that his law was 

 by no means universally applicable. It obviously can 

 only apply to cases of exclusive inheritance, and not to 

 those of blended or mixed inheritance. 



De Vries * has made similar observations to those of 

 Mendel upon varieties of no less than 15 different 

 species of plants, and in every case found that the pro- 

 portion of recessive forms obtained in the second gen- 

 eration was approximately the theoretical 25 per cent. 

 When the observations were continued through other 

 generations, the results likewise agreed with theory. 

 Tschermak f repeated Mendel's observations upon 

 the different varieties of Pisum sativum, and with 

 some of them obtained a similar result. However, 

 he found that in some other cases $ the law did 

 not hold. Correns also experimented with varie- 

 ties of the pea, and he found that whilst some of 

 the characters obeyed Mendel's law, others, such as 

 the colour of the skin of the seed, did not. He ob- 

 tained a similar result If on crossing Matthiola incana 



*Ber. d. deutsch. Botan. Gesell., xviii. p. 83, 1900. Translation 

 In J. Roy. Horticult. Soc., xxv. p. 243, 1901. 



fBer. d. deutsch. Bot. Ges., xviii. p. 232, 1900. 



JBer. d. deutsch. Bot. Ges., xix. p. 35, 1901; also Zeitschr. f. 

 d. landwr. Versuchewesen in Oesterr., iii. p. 465, 1900. 



Ber. d. deutsch. Bot. Ges., xviii. p. 158, 1900. 



IBot. Centralb., Ixxxiv. p. 97, 1900. 



