spirally Tzvisfed Races. 553 



culture grown for another purpose, I found a twisted 

 plant like the one figured and saved its seed separately; 

 its iiowers had not been protected. 



From this seed I raised in 1898 a culture of 300 plants 

 of which 259 were normal, whilst 40 exhibited torsions 

 in the main stem and 28 others in one or several of the 

 lateral branches, i. e., a proportion of 21% twisted indi- 

 viduals. I only allowed the 21 finest specimens of these 

 to flower and set seed ; and of these I only sowed the seeds 

 of the one finest twisted plant. In the following year, 

 1899, I raised from it 385 individuals of which 137 or 

 about 35% showed torsions. Here also, as in the case 

 of Dipsacns and Dianthns, many individuals exhibited a 

 ternary arrangement in the whorls. I counted about 100 

 of these, which therefore constituted about one-quarter 

 of the whole culture. 



In 1900 I grew the fourth generation, but on a smaller 

 scale. Torsions were as abundant as before, and amongst 

 their number was the plant represented in Fig. 130. 



There is little doubt that similar intermediate races 

 could be raised by an isolation of the spirally twisted 

 individuals of several other species. And the best chance 

 would obviously be given by those which frequently give 

 rise to this anomaly without selection. Thus, for in- 

 stance, Gypsopliila paniculata, Urfica nrcns (of which 

 I have already cultivated two generations with success), 

 and perhaps also Scahiosa atropurpiirca. On the other 

 hand, as stated in the previous section, my sowings of the 

 seeds of spirally twisted examples of Valeriana officinalis, 

 Saponaria officinalis, Galium Aparine^ and others, have 

 offered no such prospects. 



^Bvdragcn tot dc leer van den klenidraat. Botanisch Jaarboek, 

 Gent, IV, 1892, p. 154, PI. XV. 



