MODERN THEOKIES OF HEREDITY. 379 



The tendency of the results obtained by these biometrical 

 methods is to uphold Darwin's view of the importance of minute 

 continuous fluctuations in heredity, while the theory to the con- 

 sideration of which we will now proceed pays greater attention 

 to variation of a different type. It is necessary here to explain 

 this statement more fully. It has, since Darwin's time, been 

 customary to conceive evolution as proceeding by the gradual 

 addition or subtraction of a series of minute differences, this 

 being called " continuous variation/' but of late years a large 

 amount of evidence has been brought forward, showing that in 

 many cases variations proceed by short, but quite perceptible 

 jumps, not smoothly and continuously. This type of variation 

 has been named " discontinuous variation " or " mutation." and 

 it has even been suggested that permanent hereditary changes 

 are brought about solely by means of these mutations. The 

 chief exponent of this view has been the eminent Dutch botanist 

 de Vries. who has found and cultivated a number of new species 

 of evening primrose, which have originated by mutation from 

 the one parent species Oenothera Lamarckiana. This view, 

 certainly based on some proved facts, refers primarily to the 

 way in which variations arise, and we owe to Mendel, whose 

 name is now familiar to every biologist, the key to the problem 

 of how such variations are inherited. 



Before speaking of Mendel and his work, I will describe a 

 typical experiment of the kind associated with his name. As a 

 plant familiar to everyone in this country, the Mealie may be 

 chosen to illustrate the method employed by him. The seed of a 

 mealie contains, besides the embrv^o plant, a store of nutritive 

 material called endosperm ; in some mealies this is white, in 

 others it is yellow, and this determines the colour of the seeds 

 Now if we take some pollen from a mealie which bears seeds 

 with yellow endosperm, and use it to pollinate the female flowers 

 of a mealie which would bear white seeds, we find that the cross- 

 pollinated flowers produce only yellow seeds. If the cross is 



statistical survey of the results would give very similar results to those 

 obtained by the biometrical method. It is believed Tto take a specific 

 example) that height in man is really such a complex of characters, and 

 that it is quite impossible to separate out the single units which go to 

 make up the complex. 



Probably the strongest experimental evidence of the truth of these 

 rernarks is to be found in some recent e.xperiments of Johannsen's on the 

 weights of certain kinds of seeds. Taking a definite varietv of seed, 

 and constructing a curve to show the variation in weight of ' individual 

 grains, he finds a fairly constant '" mode " for the seed weight. Never- 

 theless, by selecting definite individuals and breeding from these, he 

 found that the " pure races " thus raised had each their own modes for 

 the seed weight, and that this did not agree with that of the whole 

 variety as a rule. These remarks will perhaps serve to explain why, 

 although both the facts and the mathematics may be quite correct, the 

 biometrical results are thought to afford verj' little insight into the in- 

 heritance of characters as compared with the Mendelian method about 

 to be described. 



