66 SCIENCE PROGRESS 



In this relation some experiments described in Dobner and 

 Nobbe's already cited work are of interest. Nobbe immersed 

 800 seeds of Robinia Pseudacacia in distilled water and observed 

 the number of seeds that swelled in each of the successive years 

 up to seven. He found that at the end of the time there were 134 

 still unswollen, but that these were living and germinated when 

 the seed-coat was injured. This not only demonstrates that 

 lack of germination of samples of seeds is often due to lack of 

 penetration of water, not to death, but also affords a means of 

 comparing the longevity and the impermeability of the seed-coat 

 of seeds. For if death be indirectly due to the permeability of 

 the seed-coat, then the seeds that swelled in each year would 

 have been the ones to die if water had reached them (from the 

 air, for instance) but not in sufficient amount to permit of growth. 

 Hence the curve denoting the number of- seeds unswollen after 

 the increasing numbers of years ought to be of a form similar to 

 that recording the number of seeds germinating after they had 

 been kept in air during increasing numbers of years. The curve 

 denoting the percentage of unswollen seeds in Robinia does 

 conform with this anticipation, as may be seen by constructing 

 it on the basis of statistics supplied by Dobner and Nobbe, which 

 are as follows in percentages : seeds not swollen after ten days 

 but swelling after one to seven years respectively, 55, 18, 7, 8, 4, 

 3, 3, 1. In this sample of 400 seeds, seventy-one swelled within 

 the first ten days, while sixty-seven remained unswollen after 

 seven years. It would, however, be more instructive to have 

 statistical work done on the longevity and permeability of the 

 seed-coat of one and the same kind of seed : with such work 

 I am unacquainted. 



The rhythmic morphological activity of phanerogams is 

 expressed in the succession of different kinds of leaves not only 

 in the life of the annual individual, but also in the growth of the 

 perennial during one season, when foliage is succeeded by scales, 

 and by flower and fruit. This order of succession is, however, 

 liable to exception, as, for instance, in those species that burst into 

 blossom before they unfold their leaves. The fact that different 

 species occupying the same habitat produce leaves and flowers 

 at different but more or less characteristic times is a sufficient 

 indication that this periodicity is largely innate. Yet external 

 factors can affect the rhythm by modifying the times of com- 

 mencement and conclusion and the duration of the cycle, so that 



