328 THE BIOLOGY OF FLOWERING PLANTS 



these adventitious embryos from the tissues of the sporophyte 

 must be regarded rather as a case of vegetative multipHca- 

 tion than as a modification of the ordinary reproductive 

 process, for neither spore nor egg cell is concerned. 

 For details the monographs of Winkler (1920) and Ernst 

 (1918) should be consulted. 



Origin of Parthenogenesis. — In parthenogenesis, apogamy 

 and nucellar budding, we might be tempted to see a new 

 evolutionary departure. That these methods have arisen 

 recently in the angiosperms is indicated by the fact that 

 neighbouring species of the same genus, or even other 

 individuals of the same species, show the normal mode, 

 but as they are found in algae and ferns, it is clear that the 

 possibility of such modes of reproduction has existed 

 throughout the history of the plant kingdom. There is, 

 however, good reason to believe that these processes are 

 abnormal. 



It is well known that in crosses between species which 

 can produce healthy hybrid offspring, these are often more 

 or less completely sterile. This sterility is due, in the 

 higher plants, to failure to form normal spores ; in many 

 cases irregularities occur in the reduction division. The 

 amount of sterility varies ; it may be complete, or some 

 good seed may be produced. So pronounced is the tendency 

 to produce imperfect spores that the presence of a proportion 

 of bad pollen, shrivelled grains, has been used as a means of 

 identifying specific hybrids in nature. Ernst (19 18) claims 

 that it is just in such natural specific hybrids that apogamy 

 and parthenogenesis are found. The abnormal relations 

 which induce parthenogenesis are due to the irregularities 

 in cell and nuclear division which occur in hybrid races. 

 Ernst supports his theory with a formidable mass of evidence 

 and it seems to fit the case in the plant kingdom remarkably 

 well. Winkler (1920) has criticised it chiefly on the ground 

 that it does not explain similar phenomena among animals. 

 Ernst's theory receives support, and a physical explanation, 

 through recent work of Haberlandt (1922). He refers the 

 stimulus which causes renewed embryonic growth near 



