48o STUDIES IN SEEDS AND FRUITS 



The seeds of J denanthera pavonina zccomplish the whole shrinking and 

 hardening process in a period not exceeding six weeks, after which they 

 assume their normal permanent condition as non-hygroscopic seeds. So 

 also with the large seeds of Entada scandens^ which within six or seven 

 weeks are transformed from soft flabby white seeds into hardened 

 mahogany-brown resting seeds that display no hygroscopic reaction. 

 The seeds of Guilandina bonducella occupy about a month in passing from 

 the soft swollen condition into the typical resting state, when they 

 respond as little as a quartz-pebble to the variations in the humidity of 

 the atmosphere. In the case of a large number of other seeds, mostly 

 permeable, a stable weight was reached in a few weeks, sometimes 

 only one or two weeks, sometimes five or six weeks, the subsequent 

 changes in weight being merely hygroscopic in character. 



However, in all these cases the experiments were made on seeds 

 detached from the parent, so that the natural process was not followed. 

 Then, again, drying in a room would probably be accomplished in half 

 the time required under open-air conditions. For instance, Horse- 

 chestnuts [Msculus) and Acorns [Quercns\ the last behaving like seeds, 

 complete their drying in five or six weeks in a room, but they would 

 probably need two or three months in the woods. In the case of 

 drupaceous fruits with a stony endocarp quite another regime is involved, 

 a matter discussed in the case of Prunus in Note 28 (p. 501). If the 

 skin remains entire several months would be needed. 



NOTE 13 (p. 446). 



Effects of very dry conditions on the growth of Bidens cernua and 

 Bidens tripartita. 



These experiments were carried out in my greenhouse during 

 three years, with the object of determining the efi^ect of very dry, 

 warm conditions on the growth of these plants. The soil was kept as 

 dry as possible, the quantity of water supplied being reduced to a 

 minimum. Under such conditions the seedhngs in the early part of 

 the experiments frequently lay prostrate on the parched soil, a few 

 drops of water soon causing them to become erect again. After they 

 were more than 2 inches high they acquired the permanently erect 

 habit. Ordinary garden soil was employed in the first year, sandy soil 

 mixed with fibre in the second year, and sandy soil alone in the third 

 year. After three generations the average height of the plants was 

 reduced from 17 or 18 inches to 5 or 6 inches. The fleshy moist- 

 tissued stems became in the first year dry and woody, in the next year 

 slender and wiry, and in the third year attenuated and thread-like. 



