BLASTOGENIC VARIATIONS. 161 



According to Darwin, variability is especially induced 

 if mongrels are repeatedly crossed with either pure 

 parent form, whilst the crossing of different species 

 may lead to much wider variation than the crossing of 

 varieties. The hybrids produced on the first cross are, 

 as a rule, fairly constant in their characters, but if these 

 hybrids be crossed again, or crossed with either pure 

 parent form, then a very considerable variability may 

 result. " He who wishes," says Kolreuter, " to obtain 

 an endless number of varieties from hybrids, should 

 cross and recross them/ 7 * Again Darwin t says that 

 cross-bred animals " for breeding are found utterly 

 useless; for though they may themselves be uniform in 

 character, they yield during many generations aston- 

 ishingly diversified offspring." Indeed it would seem 

 that entirely new characters may be produced by this 

 means. For instance, Kolreuter says that hybrids in 

 the genus Mirabilis vary almost infinitely, and he de- 

 scribes new and singular characters in the seeds, an- 

 thers, and cotyledons. Professor Lecoq also asserts 

 that many of the hybrids from Mirabilis jalapa and 

 multiftora might easily be mistaken for distinct species. 

 Again, Herbert $ has described certain hybrid Rhodo- 

 dendrons as being unlike all others in foliage, just as if 

 they were a separate species. 



According to Focke, the hybrid may be related to 

 the parent forms in three different ways: (1) there may 



* " Animals and Plants," ii. p. 254. 

 \L. c., ii. p. 74. 



\ Science Progress, vol. vii. p. 185, 1898. 



"Die Pflanzen-Mischlingen," Berlin, 1881. Quoted by Weis- 

 mann, " Germ Plasm," p. 261. 



