626 EXPERIMENT STATION RECORD. 



The Mutation Theory and Animal Breeding. — At the conference of the American 

 Society of Naturalists, held at Philadelphia during the meeting of the American 

 Association, Prof. W. E. Castle presented a paper on The Mutation Theory of < Organic 

 Evolution from the Standpoint of Animal Breeding. He stated at the outset that 

 the mutation theory was not designed to replace Darwin's theory of natural selection, 

 nor was it capable of doing so. Natural selection must still be invoked to choose 

 between different organic forms, preserving the more efficient and destroying the 

 less efficient. The question raised by this new theory is, What sort of forms are 

 subjected to the action of natural selection? Is there a complete gradation of forms 

 between two extreme conditions; and is natural selection called upon to choose from 

 the whole series the one which is organically most efficient, or is the choice made 

 merely between two widely separated conditions of ideal series? Have the varia- 

 tions in size within a species diverged by gradual cumulation of minute differences 

 in size, or by a single step? These alternative views are known, respectively, as the 

 selection and the mutation theory. 



It was pointed out that Darwin recognized no essential differences between breeds 

 and species, and believed that from a knowledge of how breeds originated much 

 may be inferred as to the origin of the species. To this end he made extensive 

 studies of breeds of animals, as well as of plants. What we need to know, the 

 speaker said, is how precisely are new breeds formed. Although they are formed 

 under our very eyes, the method eludes us. "The successful practical breeder, the 

 man who originates breeds, is a keen observer, a man of unusual intelligence and 

 skill and of infinite patience. Yet if we a<k him how in general he does his work 

 or how a particular result was obtained, we rarely get a satisfactory answer. While 

 he often withholds this information for commercial reasons, more often it is because 

 the breeder himself does not know how the result was attained. The record of his 

 breeding gives little information as to the real nature of the material used and the 

 processes involved in the formation of the breed." 



For the biologist to solve the problems involved in the formation of breeds, it was 

 urged that he must himself turn breeder and see new organic forms arise out of 

 material which he is thoroughly familiar with, and under conditions which he can 

 control. Little work of this sort has yet been done, and generalizations can as yet be 

 made only tentatively. So far as the evidence goes, it was said to indicate that the 

 material used by breeders for the formation of new breeds consists almost exclusively 

 of mutations. "The breeder does not set to work with some purely imaginary form 

 in mind toward which he seeks by selection gradually to mold his material. He 

 commonly either discovers the new breed already created and represented by one or 

 more exceptional individuals among his flock, or else he seeks by crossbreeding to 

 combine in a single race characters which he finds already existing separately in 

 different races. In both cases he deals with mutations, that is, with characters 

 unconnected by a series of transmitted stages with the normal form." 



Illustrations were given from the author's own experience in breeding guinea pigs. 

 In three generations lie was able to establish a race with a well-developed fourth toe 

 on either hind foot. This was not created by selection, but was improved by that 

 means; it was born and not made — a mutation. Similar results were observed in 

 breeding long and short haired guineas. If these results had taken place in nature, 

 natural selection would have followed to determine which was best suited, and the 

 different breeds would have been geographically separated. 



The combining in one race of the characters already found in different races, by 

 means of crossbreeding, was shown to depend upon the facts that (1) mutations are 

 alternative in heredity to the normal condition, and (2) one mutation is entirely 

 independent of another in heredity. From the array of forms obtained, the breeder 

 can select the particular combination of characters which suits his purpose. Cross- 



