390 W. Bateson. 



is sufficient to produce pigment. The pbenomena in the animals 

 prove either that 



(1) the pigmentation depends on one factor only; 

 or (2) that the two factors have not been separated in variation ; 

 or (3) that one of these two factors is always present in every 

 animal. 



Between these alternatives it is as yet impossible to decide, but 

 periiaps the third is the more probable. 



Interdependence of Structure and Colour. 



The most elaborate analysis of this kind yet made is that 

 resulting from Miss Saunders' experiments with MaUhiola. In 

 the case of Ten-week Stocks she has proved that sapcolour in the 

 flower is caused by the meeting- of two factors just as in Lathynis 

 odoratns. But in addition to this, the development of the hoariness 

 or felting of hairs upon the surface of the leaves is also produced by 

 a third factor which may be present alone without causing a single 

 hair to be formed. The results proved that though the factors for 

 hoariness may be present, the zygote is glabrous unless it also 

 possesses flower-colour. For the development of hoariness three factors 

 are thus necessary, including the two which cause sap-colour to appear. 

 (In the Brompton Stocks hoariness may however be formed in the 

 absence of sap-colour, and the statement given is only applicable to 

 the Ten-week Stocks.) 



As may be supposed the existence of these specific peculiarities 

 made the unravelling of the Stock problem extraordinarily difficult, 

 and as similar cases will probably be met with hereafter some of 

 the steps in the argument may be recapitulated with advantage. 

 (See 10 b.) 



When hoary plants (H) were crossed with glabrous (G), hoary is 

 dominant, and Fo consists of 3 H : 1 G. Naturally it was to be ex- 

 pected that Fj bred with a glabrous would give e(iual numbers of 

 H and G. It was found that when the original G was used, this 

 simple result followed, but when G plants with diff'erent flower-colours 

 Avere used, the offspring were all H. 



This fact led to an investigation of the inter-relationships of the 

 glabrous varieties, when it was found that certain glabrous strains, 

 though breeding true on selffertilisation, give hoary plants when 

 crossed together. This therefore Avas reversion in a structural 

 character occurrins- as the result of a cross. The following table of 



