166 ON BIASTREPSIS IN ITS RELATION TO CULTIVATION. 



Sowing of July 28, 1892. Seeds harvested in 1891 were sown on 

 a bed having an area of 4 square metres, and the soil was kept moist 

 by means of a lawn-sprinkler. The seeds germinated rapidly and 

 before the autumn had formed vigorous rosettes of radical leaves. 

 By the end of the following May all the seedlings, with the exception 

 of a few weakly ones, had thrown up their shoots. When the shoots 

 had attained a length of 50-75 cm. and the terminal capitulum 

 could just be seen, it was found that all the plants were quite normal, 

 having decussate phyllotaxis; the only observable abnormality was 

 one divided leaf. There were 131 plants with elongated shoots and 

 twenty-nine with only rosettes, 160 plants in all, on the 4 square 

 metres, so that the proportion (forty) per square metre was high; 

 but this number did not appear to be excessive, inasmuch as the 

 plants were relatively small and did not touch each other more than 

 would vigorous plants at a greater distance from each other. After 

 the counting was done, I allowed five of the best specimens to remain ; 

 they developed strong stems over 2 metres high, which were not 

 noticeably less vigorous than ordinary normal plants of the breed. 



In a control-experiment, made with the same seed sown in the 

 spring of 1892, I obtained twenty- two individuals out of sixty, that 

 is, almost 37 per cent., having twisted stems. 



Sowing of August 14, 1893. The seed used was obtained from two 

 plants of the 1892 crop and was the same as that used in the June 

 sowing of this year. They quickly germinated in the bed and by 

 the autumn had formed relatively small rosettes, so small, in fact, 

 that sixty plants per square metre could well be left. Notwithstanding 

 this, they nearly all (235) threw up their shoots in the following 

 spring; but here and there, where germination had been tardy, some 

 plants, twenty in all, remained as rosettes, that is, about 8 per cent. 

 The stems were weak, of only about half the normal thickness, and 

 their leaves were decussate up to the inflorescence; there was no 

 sign of twisting of the stem. 



Summarizing these results of summer-sowing, we find that, given 

 adequate space and suitable treatment, there is an almost entire 

 absence of twisted stems. The experiments included 179, 131 and 

 235 plants, giving a total of 545 shoots, of which only one was 

 twisted and one had threeleaved whorls, both of these exceptional 

 plants belonging to the June sowing. In contrast with this, the 

 spring-sowings gave 34-37 per cent, of twisted stems. 



I conclude, therefore, that late sowing does not allow the plant 

 sufficient time in which to attain normal vigour in the rosette-stage. 



