78 THE EVOLUTION OF LIVING BEINGS. 



inadvertently crossed to the extent of i % by other 

 forms, RiMPAU found similar numbers in the case of 

 four-rowed barley, and Mayer-Gmelin even higher 

 ones in beans. 



Yet, in a very general way, the difference in behaviour 

 between selffertUizers and individuals, mating at ran- 

 dom after a cross, is correct. 



Now it is a fact, fully worth our attention, that — in 

 a general way also — what we find in nature, agrees with 

 these calculations, as far as selff ertilisers are concerned, 

 but, as a rule, does not agree when random-mating 

 occurs. 



Selffertilizing Linneons consist really of a great 

 number of distinct forms — though generally not so 

 astonishingly pure as the calculation would make us 

 suppose — while Linneons, within which random-ma- 

 ting occurs, usually have a rather uniform aspect. 



Of course there is no sharp hmit between these two 

 categories of Linneons, and this caimot be expected 

 either because strict self fertilisation occurs as little as 

 absolutely promiscuous mating, and what seems, at 

 first sight to be luiiform, proves sometimes to be mul- 

 tiform on closer examination. 



So f . i. R. E. Lloyd (The Growth of Groups in the 

 Animal Kingdom. Longman Green & C°. 1912. 185 

 pp.) fide American NaturaUst 47. 1913, on studying 

 rats in connexion with the plague problem in India, 

 found that small groups of rats, differing in some res- 

 pects from the forms regarded as tj^pical, occur here 

 and there. But even if we take full account of this lack 

 of a sharp hmit between self-fertilisers and random- 



