30 THE EVOLUTION OF LIVING BEINGS. 



conditions when all possibility of a cross is excluded. 



This is not inconceivable; it could be brought about 

 in two ways: certain gametes might loose a property, 

 they previously possessed in common with the other 

 gametes, or some gametes might gain a property which 

 they, as well as the other gametes, previously lacked. 



Such a processs is conceivable, but it would add so- 

 mething fundamentally new to our stock of knowledge ; 

 something entirely out of the pale of Mendelism f. i. 

 and to which therefore Mendelian behaviour could give 

 no direct support, although the contrary is believed so 

 frequently, because Mendelism deals with the demon- 

 strated fact that a heterozygous individual produces 

 more than one, a homoygous individual but one kind 

 of gametes,and has nothing to do with the way in which 

 heterozygotism arises, except if it does so by crossing. 



But the variability required by evolutionary theories, 

 based on a process of variation, should be able 

 to change a homozygous organism,z&tV^OM^crossing,into 

 a heterozygous one. 



We repeat : this is not inconceivable, but very clear 

 proof for its real existence is wanted, before we are 

 justified to accept it, as sufficient evidence for such an 

 unexpected and novel occurrence. 



Now how can we possibly get this proof? 



Mendelism can 't give us the clue; it does not deal 

 with the internal structure of the gametes, except in so far 

 as deductions are concerned, derived from the/ac/ that 

 the hybrid arisen from the product of the fusion of two 

 gametes of different constitutions, is able to form more 

 than one kind of gametes. 



