248 ESSENTIALS OF BIOLOGY 



controlled breedings and selection and, of course, it may occur, 

 of itself, in nature but only with great improbability. Still we 

 recognize the Mendelian category as a possible one, more or less 

 probable if there is some agency in wild nature that corresponds 

 to the agency represented by the experimentalist, or breeder in 

 the laboratory or farm. By selection and inbreeding, then, we 

 can make a category of organisms that have always the same 

 characters (apart from fluctuations) from generation to generation 

 — that " breed true." 



It is also possible to establish " pure races " that are produced 

 by asexual, or parthenogenetic reproduction. That is, it may 

 be possible to select individuals from the local race that display 

 some particular variation and then to reproduce these individuals 

 asexually (for instance, by grafts, cuttings, etc., in the cases of 

 plants). It follows, of course, that we must find inheritable 

 variations, by trials. Then we rear a series of generations, 

 from one ancestral organism, that diifers somehow from other 

 series of generations reared from other ancestral organisms. 

 These are " pure races " produced, as in the cases of Mendelian 

 categories, by selective breedings. They are what we may call 

 irreducible organic categories. Their characters can be defined by 

 some description, or diagnosis. We breed the individuals of the 

 category among themselves and we observe that, from generation 

 to generation, they " breed true," that is, in every progeny, or 

 progeny of a progeny, the characters, as stated in the diagnosis, 

 are reproduced. The diagnosis, or definition of the category 

 will be, of course, wide enough to include those slight variations 

 from " the ordinary " that we call fluctuations. That such 

 fluctuating variations are " accidental " or non-essential to the 

 diagnosis we prove by attempting to reproduce them. That is, 

 we may select individuals from an irreducible category that dis- 

 play some exceptional value of a character (say great size of body) 

 and then breed these individuals among themselves. We find, 

 then, that the progeny, or the progeny of the progeny, always 

 revert back to the original ordinary characters. 



If then we study some irreducible category and find among 

 its individuals some which do not conform to the definition, and 

 have progeny which also do not conform to the definition, then 

 we have observed transformism to occur. In order that such 

 transformism may involve evolutionary change it is necessary 



