546 Heritable Spiral Torsions. 



time familiar with the most favorable conditions of this 

 culture, and obtained only two twisted individuals amongst 

 nearly 1650, probably for this reason alone. These two 

 flowered in isolation and set an abundance of seed. The 

 third generation (1888-1889) raised from seeds of the 

 second, produced 67 twisted stems in a culture of about 

 the same extent, i. e., about 4%. Here again the seed- 

 parents flowered in isolation. 



The fourth generation was sown partly in 1890, and 

 produced as many as 10% twisted individuals which 

 however could not be used for the continuation of the 

 race. The rest of the seed was sown in 1891 with a 

 better knowledge of the conditions, whereby the propor- 

 tion of twisted individnals rose to 34%, a value which 

 the later generations on the average maintained, without 

 surpassing it to any considerable extent. 



The improvement in the method of culture consisted 

 essentially in providing the young plants from their very 

 earliest stages with more room. In the two previous 

 generations about 50 specimens were grown per square 

 meter In this generation, however, the number was 

 reduced to about 25 by the removal of the superfluous 

 ones in June, as soon as the ]:)lants began to touch one 

 another. 



The seed had been sown in May in the beds. At the 

 beginning of October I noticed, amongst about 100 indi- 

 viduals, 6 with a s])iral arrangement of leaves in the 

 heart of a rosette. In the beginning of November I was 

 able to remove more than half of the plants as undoubted 

 atavists, and when, towards the end of May 1892, the 

 stems grew up vigorously, they were finally examined 

 and recorded. The result, together with that obtained in 

 November, i. e., for the whole culture, is as follows: 



