8o Saturated and Dihite Colours [CH. 



The FZ ratio 9:3:4, the significance of which we 

 have been considering, is one which very frequently recurs 

 in Mendelian analysis. For example, as Tschermak found, 

 when a pink-and-white flowered eating pea (Pisum sativum] 

 is crossed with a white flowered type, F^ is often of the 

 original purple flowered type. Then F. 2 will be 



9 purple : 3 pink-and-white : 4 white. 



Similarly pink Salvia Horminum x white may give F l 

 purple, and F z 9 purple : 3 pink : 4 white. In these cases 

 the factor for the purpleness is of course brought in by the 

 albino, but exactly the same F z may result from a cross 

 between the purple type and an albino not carrying the 

 factor for purpleness. All that is essential for the produc- 

 tion of this ratio in F z is that F^ should be heterozygous 

 for two factors, of which one is perceptible whenever 

 present, while the other needs the presence of the first in 

 order that its own effects may be manifested. Such cases 

 are very numerous and in practical breeding are to be 

 looked out for continually. Care must be taken to distin- 

 guish them from families like those of the Andalusian fowl 

 (p. 52) in which the commonest term in the F z series is a 

 heterozygous type. There the numbers will be 1:2:1, 

 which in a practical example may give results not obviously 

 distinguishable from 3 : 9 : 4. To decide between the two 

 possibilities it is necessary to breed the F z types again. If 

 neither of the scarcer types when bred inter se can throw 

 the other, and the commoner type cannot be bred pure, the 

 latter is a heterozygous type; but if one of the scarcer types 

 can throw the other, then the ratio is presumably 9:3:4, 

 and in such a case it will be possible to raise a breed true 

 to the type occurring as 9. 



Saturation and Dilution of Colours. 



Omitting yellow from our consideration, we thus re- 

 cognize that in the mouse the colours, grey (agouti), black, 

 chocolate, which the fur visibly presents, result from the 

 interaction of several factors, and that these factors can in 

 great measure be shown to be distributed in gametogenesis 

 according to Mendelian allelomorphic systems. The ex- 



