BRIEFER ARTICLES. ^ 



SUGGESTIONS ON THE CLASSIFICATION OF SEEDS. 



Modern botanical text-books do not present a classification of 

 seeds in accordance with modern ideas of their structure and function. 

 It comes as a distinct surprise to the average student of plants when 



r 



he, for the first time, appreciates that some familiar herb, shrub, or tree 

 may bear during a single sep.son a score or more different types of 

 seeds. Nor does the great complexity underlying the concept of the 

 seed seem to be fully comprehended by the plant breeders and seed 

 growers. Indeed, no method has been devised for recognizing the 

 types of seeds that theoretically exist, nor, so far as I am informed, has 

 a nomenclature been provided by which one type may be delimited 

 from another. Under these circumstances it seems desirable to set 

 down a simple classification of the more fundamental types of seeds 

 in the hope that future experiment will indicate marks by which they 

 may be recognized, and that some new light may thus be thrown upon 

 the problems of hybridization, seed control, variation, mutation, and 

 evolution of varieties and species. 



To define the word '*seed'' in a manner technically accurate is 

 exceedingly difficult. Underneath the seed lie the following con- 

 ditions, in the absence of any of which the seed does not exist as 

 such : 



1. Alternation of generations. No seed is ever produced by non- 

 alternating types of plants. 



2. Heterospory. Among alternating types seeds are not produced 

 by homosporous forms. 



3. Intraspecificsymbiose. Heterosporous types devoid of intergen- 

 erational symbiosis do not produce seeds. 



The seed is in no exact sense an organ of the plant that bears it ; 

 it is rather an aggregate in the family life of the species. In it tis- 



t 



sues and structures belonging to three consecutive generations are uni- 

 formly present either transitorily or permanently. While its significance 

 for the persistence of the species may be very great, it cannot be 



224 [SEPTEMBER 



% 



^ 



i 



r 



rv 



I 



i 



" 



