i88 



STUDIES IN SEEDS AND FRUITS 



The great 

 range in the 

 proportional 

 weights of 

 the seed's 

 coverings in 

 the resting 

 state. 



them with the same data for other seeds, but more especially 

 when we compare them with the proportional weights in the 

 pre-resting seed of the moist fruits, and in the resting seed 

 swollen for germination. I here append in a tabulated form 

 my results stated as percentages for about eighty-two species of 

 plants, arranged in order, the seeds with the smallest proportions 

 of coverings coming first. The data are given in two tables, 

 one for leguminous seeds, and the other for seeds of other 

 orders. For convenience only one value will be referred to 

 in this discussion, that for the seed's coverings, which I will 

 term the " seed-coat ratio." Thus, when Iris Pseudacorus is 

 said to have a seed-coat ratio of 20, it is meant that the 

 coverings make up 20 per cent, of the weight of the entire 

 seed. 



We notice in these tables that there is a great range in 

 the proportional weight of the seed's coverings, namely, 

 from 5 to 69 per cent. The maximum of the range is but 

 slightly extended in the case of seeds with abundant hairs. 

 If we were to exclude them altogether we should still 

 have the hairless seeds of Sapindus Saponaria representing 

 the maximum limit of 64. But we should be no more 

 justified in disregarding the hairs of a seed than we should 

 in leaving out of our calculation the wing-like appendages 

 of seeds like those of the Mahogany tree (Swietenia) or of 

 the Moringa tree. It is noticeable that whether we restrict 

 ourselves to leguminous seeds or to seeds of other orders, 

 the range is much the same. The leguminous seeds belong 

 to 39 species and show a variation from 6 to 61 per cent. 

 The other seeds, which belong to 43 species of 20 different 

 orders, give a range from 5 to 69 per cent. Since, therefore, 

 nearly the whole range is to be found with leguminous 

 seeds, it follows that with seeds of this order there is nothing 

 distinctive in the proportional weight of the seed's cover- 

 ings ; and, generally speaking, except in the case of seeds 

 having a ratio less than 20 per cent., they run fairly well 

 together. 



