1 96 LOCK : 



writer to Ceylon at the beginning of 1908 ; and the F 2 gene- 

 ration was raised at the Peradeniya Botanic Gardens from 

 this seed. 



Most of the more important families, which should have 

 formed the third generation of hybrids, have failed to germi- 

 nate ; and, as it is doubtful how soon the experiments can be 

 repeated, it was thought desirable to publish a brief sketch of 

 the results so far obtained. No one is likely to be more 

 conscious than the writer of the incompleteness of these results, 

 but it is hoped that the experience gained so far will be of use 

 to other workers. 



The object with which these experiments were begun was 

 in order to throw light, if possible, upon the question whether 

 characters are to be found among plants of such a kind that 

 their inheritance cannot be brought into line with the results 

 of Mendel ian investigation. This is a subject which has 

 recently come in for a good deal of discussion. It has been 

 suggested in several quarters that the Mendelian phenomena 

 are confined to cases of crossing between types which owe 

 their origin to the effects of cultivation.* One critic has even 

 gone so far as to describe the Mendelian rules as representing 

 pathological phenomena confined to the unnatural forms 

 produced by man's interference with certain types of life. 

 The behaviour, it is asserted, of true natural species, when 

 crossed together, is entirely different. 



Those who claim in this way that a difference in kind exists 

 between variety crosses and species hybrids rely chiefly upon 

 the authority of de Vries ; and they frequently forget that 

 the distinction drawn by de Vries was not between individuals 

 ,as units, but between different kinds of unit characters. De 

 Vries in fact maintains that two types may differ from one 

 another bv segregating (Mendelian) characters, or by non- 

 segregating characters, or by both. On the other hand, a 

 considerable number of students of heredity, impressed by the 

 steadily increasing number of cases in which all kinds of 

 characters can be shown to conform to the Mendelian rules, 

 await a more certain proof of the existence of true blended 



* Although it has not been satisfaritorily shown that simple culti- 

 vation can cause the origin of anything new. 



