22 



food, light, warmth, and other external conditions, and the lat- 

 ter from the energy stored in the seed, it is necessary to bring for- 

 ward still other data. We may venture to formulate this proposed 

 extension of the law relating to the size of the seed thus: large seeds 

 give rise to plants with a greater a evolopment of the reproductive parts 

 and less of vegetative parts , than small seeds do. 



It is intended here to directly compare the reciprocal relations of the 

 two sides of the plant as influenced by the parent seeds. The data may 

 be taken by weighing the fruiiing portion and comparing it with the 

 weight of all the remainder of the plant, both done when at their best 

 development ; or other methods may be used. 



Excellent data are supplied from the researches of Lehmann (see 

 table VIII). He grew large, medium and small peas, over 400 of 

 each lot, and obtained plants that were heavier for the larger seed in 

 both their vegetatix e and their reproductive parts, i.e., the leaves and 

 stalks for the vegetative part, and the peas and pods for the reproduc- 

 tive part. A nd yet when ihe weight of the vegetative portion is com- 

 pared with that of the reproductive portion of each lot, it is clear that 

 the fruiting pait has attained a stronger development in comparison 

 to the remainder of the plant in the lots from larger seeds. To state 

 the facts in another way, the larger seeds not only grow larger plants, 

 but those which have fruiting parts more strongly developed th^n the 

 associated vegetal ive parts. 



Interesting data are furnished by Birner ami Troschke using oats 

 and peas, and by Marek with peas. The last investigator found that 

 the weight of peas of first quality was nearly three- fourths of the 

 whole harvest raised from large seeds, and only about one third of that 

 from small seeds. (See table VII). In this case, therefore, the large 

 steds not only gave a much better total yield, but far more seed ma- 

 terial of high grade with which to continue the strain. 



Marek, in Germany, experimenting with wheat (see table IX), and 

 Plumb in the United States, with oats (see table X), have demon- 

 strated the same fact. Both have provided data which show that the 

 amount of grain in comparison with the straw was greater in case of 

 large seeds ihan of 6mall ones. 



Statistical evidence oft his kind might be greatly extended although 

 observations have rarely, if ever, been instituted with this particular 

 end in view. Casual observations give no aid to this part of the en- 

 quiry, bs the differences are obscured by other factors which stand out 

 more prominently. What the eye cannot detect, however, is readily 

 and unmistakeably revealed by the rule and balance. 



So far as data can be marshalled at present, there appears good rea- 

 son to believe that large 6eeds, besides giving rise to larger and more 

 fruitful plants, also possess an inherent tendency to accentuate the re- 

 productive side ( f the resulting deve'opment. If peas are sown, the 

 largest seeds not only give rise to the largest plants, with the great- 

 est weight of pods and of seeds, but to an excess of fruitage when 

 compared with the remainder of the plant ; and in a similar way with, 

 other kinds of plants, the largest parent seeds give the greatest re- 

 turns of fruit and daughter seeds, both absolutely and also in compar- 

 ison with the growth of leaf, stem and root. It is to be understood^ 

 of course, that we are not attempting to deal with single plants, but 



