90 Menddicw Tulurltance and Ytllov Riixf in Whtat 



that fresh coiiiljinations of those features which more especially alTect 

 the plants laetabohsm may also modify its power of resistance. The 

 parents of the cross under consideration were dissimilar in respect of 

 several such characters, e.g. fohage, period of ripening, and probably 

 also in root-range. As 13iflen(3) has pointed out, when segregation occurs 

 fresh combinations of these various features result, and consequently 

 the metabolism of such offspring is affected in various ways. The resulting 

 root-range may be disproportionate to the new leaf-area, a slower rate 

 of maturation prolongs the period of possible attack, and so on. Most 

 important of all, it must not be overlooked that physiological features 

 are also inherited, and recombinations of these may produce unexpected 

 results. 



While some of these new combinations may bring about increased 

 susceptibility, there was evidence that others may reduce it. or stabilise 

 the inherited resistance. This fact has also been independently noted by 

 Hayes (8) in his experiments. These effects are realized not only in the 

 Fo, but also in any subsequent generation in which such combinations 

 of features are iidierited, or arise. A consideration of such points as 

 these enables one to understand partly, at least, why statistics of rust 

 attack on an i^g generation (from a cross between susceptible and im- 

 nume parents) seldom give a close approximation to the 1:2:1 

 Mendelian ratio. 



{(l) Variation of .susceptibility in hybrid wheats. 



The view that the re-combination of parental characters may lead 

 to a modification of a plant's predisposition to attack is further supported 

 by the various degrees of susceptibihty posse.ssed by hybrids. It has 

 been stated by several observers tliat hybrid wheats are usually as 

 susceptible to rust as the more susceptible jiarent. Instances, however, 

 are known to the writer in which the attack was distinctly more severe 

 on the hybrid, and also other cases in which it was much less severe. 



In order to make a strict eom])arison on this point, certain liybrids 

 were grown alongside their parents in 1911) under exactly the same con- 

 ditions as to time of sowing, space allowance, etc. One of these crosses 

 was between American Club and a very susceptible race No. 109/1. Of 

 the 14 hybrids, none had more than a slight attack at the final examina- 

 tion on July 31st, while all the plants of 109/1 were badly rusted. 



Another cro.ss was made between a moderately susceptible variety 

 (Brooker's) and American Club. Sixty-eight t\ plants were raised in 



