A. St Clair Caporn 239 



The first thing to notice is that the pure tights make up ahnost 

 exactly one-quarter of the total. The agreement with expectation is 

 remarkably good. 



The tight-containers call for closer study, as their group is really a 

 mixture of two. These gi'oups can only be defined and separated by 

 further breeding. For whereas the one type always throws pure tights, 

 the other, which is outwardly indistinguishable from it, never does so. 

 In some cases as many as 35 — 65 plants and more have been raised 

 from tight-containers of this second kind without any pure tights 

 appearing among them. It was also found that out of a fair random 

 sample of 118 tight-containers 7 behaved in this way. 7 out of 118 is 

 equal to 78 out of 1310; so that the results in Table IV may now 

 be summarised as follows : 



78 Tight-containers not 

 throwing pure tights. 

 610 Pure tights : 1232 Tight-containers: i 369 Hardbacks. 



throwing pure tights jHO Penulti-loose.s. 

 46 Pure looses. 



603 



This definitely establishes the 1:2:1 ratio, or, in other words, the 

 determination of complete tightness by a single independent factor. 



But what is it that causes the heterogeneity of the last term in 

 the ratio, the G03 plants which, while so obviously of difierent sorts, 

 have the one common property of being unable to give rise to pure 

 tights ? The 46 pure looses compose roughly one-sixteenth of the 

 group. Added to the penulti-looses they form one-quarter. These 

 proportions are significant : plainly other factors must be coming into 

 play. 



At the present stage of the investigation, mainly through an in- 

 sufficiency of available experimental evidence, it is not possible to give 

 the exact number and mechanism of these factors (whether they are 

 complementary or cumulative, or even inhibitory, for example), but it 

 seems tolerably certain they are there, and that only through them 

 can the very variable constitution of the F^ individuals (see Table II) 

 be accounted for. 



Let us suppose, in order to illustrate this view, that there are 

 3 factors : 



X = a factor capable of rendering all the paleae on the plant pure 

 tight. 



