E. 1{. Saunders 155 



obtained in F^. The mixed offspring from the quarter-hoary F.^ plants 

 occur in such proportions that we can hardly doubt that, had the 

 45 individuals which died early lived to maturity, the number to be 

 added from the doubtful coluum to the quarter-hoary and glabrous 

 classes respectively would have brought the totals into close agreement 

 with expectation. As was anticipated all the glabrous and half-hoary 

 individuals that were tested bred true. 



The above series of results forms a concordant whole and gives us 

 a new insight into the relations of the factors determining surface 

 character. Taken in conjunction with what we already know they are 

 explained on the following simple scheme : — 



(1) That the half-hoary type lacks H but contains K together with 

 an additional factor (indicated by J) which reacts with K to produce a 

 certain degree of hairiness but definitely less than that exhibited by 

 the fully hoary type. 



(2) That when the amount of the dose, if we may so express it, of 

 J is equal to that of K, U being absent, the result is the condition 

 termed half-hoary, as seen in the half-hoary type employed which we 

 may write CRJK. 



(3) That when, on the other hand, one dose of J goes with a 

 double dose of K, H as before being absent, as we have in the matings 

 marine {CRK) x half-hoary, white {GK) x half-hoary and sulphur-white 

 {CK) X half-hoary (see Experiments 4, 5 and 6), the result is to dilute 

 the effect of J and to produce the quarter-hoary condition. 



(By analogy it would be natural to sujjposo that in the converse case 

 where the dose of ./ is double that of K, i.e. in the combination JJKk 

 instead of JjKK, the individual would be intermediate in hairiness 

 between the half-hoary and the fully hoary state, or three-quarter- 

 hoary. The only mating among those given above which would pro- 

 duce this particular combination in F.^ is CRJK x CRH (half-hoary 

 X glabrous flesh) which gave F^ as shown in Experiment 7. As the 

 expectation in this case would only be 1 three-quarter-hoary individual 

 in 32 the fact that none were recorded among the 44 plants actually 

 raised must not be regarded as conclusive evidence that a three-quarter- 

 hoary form recognisably distinct from the half-hoary and the fully hoary 

 does not exist. For the present this point must remain undecided. 

 We may however go so far as to predict that whatever the appearance 

 of individuals of the composition CRJKCRJk they will in behaviour 

 resemble the quarter-hoary form in that they will not breed true owing 

 to the non-existence of a gamete corresponding to this form of zygote.) 



