162 ON BIASTREPSIS IN ITS RELATION TO CULTIVATION. 



I made a second experiment, in wliich the plants were still more 

 crowded together. In the first summer I allowed 136 plants to 

 grow on a bed of 2.1 square metres area, that is, sixty-five plants to 

 the square metre. In order to eliminate the effect of the border- 

 position, I surrounded the bed with a margin of plants placed about 

 as closely as the others, but these marginal plants were not subse- 

 quently counted in with the others. When I examined the rosettes 

 in the following February, no trace of spiral arrangement could be 

 detected. A control-experiment, with sixty-two plants from the 

 same seed, but planted twelve to the square metre, gave 10 per cent, 

 of individuals with twisted stems. 



Hence lack of sufficient space can entirely prevent biastrepsis 

 from making its appearance. 



A similar result was obtained with regard to the exposure of the 

 beds. Shade, whether of trees or other objects, is always prejudicial; 

 a good proportion of twisted stems can only be obtained when the 

 plants are grown in an open situation fully exposed to the sun. 

 Thus, in 1890, a sowing in the shade of a tree gave only 3 per cent., 

 whilst another quite similar sowing, but with full exposure to the 

 sun, gave 7 per cent, of plants with spiral phyllotaxis. 



The time for sowing the seed, whether in the open or in pans in 

 the greenhouse, varies from March to the beginning of May. The 

 time and mode of sowing seems to be of little or no importance as 

 compared with the requirement of sufficient space. 



When sowing in the open, about 3-5c.cm. of seed per square metre 

 should be used, containing 200-300 seeds. Of these only a portion 

 will germinate, and of these some are lost or are weeded out, so that 

 eventually only about 20-25 plants result. On the other hand, seeds 

 sown in pans kept in the greenhouse nearly all germinate ; the seedlings 

 are very uniform and can be potted off without any selection. The 

 latter method therefore yields by far the more reliablef igures of the two. 



In weeding out the seedlings obtained by sowing in the open, 

 either the most vigorous plants may be left, or no attention may 

 be paid to this point; but the result is materially influenced by the 

 course pursued, for the weaker plants are much less likely than the 

 vigorous to show twisting. They are plants which apparently have 

 been badly nourished, either in the seed-stage whilst still on the 

 parent, or during or after germination. I made an experiment on 

 this point in 1888-9. In June 1888 I planted two similar beds, 

 each of 12 square metres area; the one with 360 very vigorous plants, 

 the other with 410 weakly plants, from seeds of the same parent. 



