TJic Spiral Disposition of tlic Leaves. 535 



of normal length, for instance of a decimeter or more, 

 in the middle of a twisted stem. 



Like other anomalies spiral torsions are to a very 

 large extent dependent on external conditions. Under 

 unfavorable treatment the anomaly may be almost en- 

 tirely absent, even from culttires from seeds which other- 

 wise would gi\e rise to a third or mcn'e of individuals 

 with fine spiral torsions. The experiments which I have 

 made on this point with Dipsaciis sylvestris tarsus and 

 which have been confirmed by the results of my experi- 

 ments with other species, seem to justify the following- 

 conclusion.-^ 



The more favorable the conditions of life and the 

 more vigorous therefore the growth, the greater will be 

 the proportion of beautifully twisted plants in a given 

 culture, and the more complete will be the torsions which 

 are produced. 



The most important of these conditions is the space 

 given to the individual plants. They ought not to be 

 shaded by one another nor touch one anotlier. More 

 than 20-25 plants should never be grown on a square 

 meter. Grown thus they nevertheless stand in close con- 

 tact in autumn, and it is obviously still better that no 

 more than 10-15 plants should occupy a square meter. 



Fewer torsions are produced on a given area when 

 the plants are crowded than when the distances ])etween 

 them are greater. No useful purpose is therefore served 

 by increasing the number of plants on the same bed. Tf 

 nevertheless this is done the twisted individuals will be 

 found almost exclusively along the edges of the bed. 



The time of sowing the seed is a point of considerable 



^ O;; Biasfrcf>sis ?';; Its Relation to Cultivation. Annals of Botany, 

 Vol. XIII, 1899, p. 395. 



