REPRODUCTION BY SEED 165 



cross-pollination that they were inclined to minimize the 

 importance and extent of self-fertilization, which was 

 regarded as not merely undesirable, but, as a general rule, 

 positively harmful to the race. This is an exaggeration, 

 for in most cases self-fertilization can certainly take place 

 for many generations without impairing in any way the 

 vigour of the stock. At the same time, the importance of 

 cross-fertilization in maintaining a good average race is 

 unquestioned, and Robert Knight, in 1837, was no doubt 

 right in the main when he said that self-fertilization was 

 not possible for a perpetuity of generations without the 

 intervention of cross-fertilization. 



The truth of the matter is that self-fertilization has no 

 more harmful effect on future generations than vegetative 

 reproduction ; probably less. Vitiated habits and objec- 

 tionable characters acquired by a parent are transmitted 

 by vegetative multiplication, but not by seed, even when 

 self-fertilized, unless they have become so pronounced as 

 to impair the nutrition and check the development of 

 their possessor. The same is true of in-breeding in 

 animals. If the parents are sound, the progeny is sound. 

 The danger arises when unsoundness comes in. By in- 

 breeding this is intensified ; by cross-breeding it is modified 

 or eliminated. Again, self-pollination is economical, a 

 large number of seeds being fertilized with the minimum 

 expenditure of pollen. Moreover, the certainty of pollina- 

 tion is greater than in cross-pollinated plants, and generally 

 the output of seed in these plants is very great indeed. 

 This accounts for the wide distribution of autogamous 

 species, and the rapidity with which they couquer new 

 ground. Many of the commonest and most widespread 

 weeds are self-pollinated— e.gr., groundsel (Senecio vul- 

 garis), chickweed {Stellaria media), shepherd's-purse (Cap- 

 sella Bursa-pasforis). These all have small inconspicuous 

 flowers which rarely open. Many other weeds are only 

 occasionally cross-pollinated, and even in ordinary insect- 

 poUiaated flowers self-pollination very often steps in 

 when cross-pollination fails. The garden-pea, in spite of 

 its attractions, is invariably self-pollinated. In some 

 plants seeds are formed in flower-buds which never open. 

 These closed or deistogamous buds (Gr. cleistos, closed) 

 are found in many species of Viola, wood-sorrel (Oxalis), 

 and stitchwort. These plants are visited by few insects 



