NATURE OF MENDELIAN UNITS 



All Extensive Researches Indicate that so-called Units are Groups of Very Small 



sub-characters, and not Indivisible Units — Alternative Inheritance no 



Evidence of Discontinuous Evolution — Importance to Eugenics of 



Understanding True Nature of Characters. 



G. N. Collins 

 Bureau of Plant Industry, U. S. Department of Agriculture, Washington, D. C. 



THE discontinuous nature of Men- 

 delian inheritance is often taken 

 as evidence that evolutionary 

 progress is also discontinuous. 

 When two varieties appear to differ in a 

 single conspicuous character and this 

 character is found to behave as a unit 

 in the progeny of crosses between the 

 two forms, it seems reasonable to infer 

 that the character sprang into existence 

 abruptly. The failure to secure inter- 

 mediates among the hybrids certainly 

 suggests that there was a corresponding 

 absence of intermediates when the 

 character arose. 



This line of reasoning has not gone 

 unchallenged, but most Mendelians 

 continue to look upon characters as 

 units and many are so unbending as to 

 continue in the belief that the units are 

 unchangeable. It is the purpose of the 

 present paper to discuss the bearing of 

 Mendelian inheritance on the origin of 

 characters and to suggest an extension 

 of the theory of factored characters. 

 Reasons are advanced for believing that 

 all visible characters are complexes of 

 minute sub-characters which have be- 

 come coherent through long association. 

 During the years immediately follow- 

 ing the revival of Mendel's discovery, 

 many alternative characters were in- 

 vestigated and the crosses were regu- 

 larly reported as showing no inter- 

 mediates and as giving in the second 

 generation the 3 to 1 ratio of the simple 

 Mendelian hybrid. Results of this kind 

 are most easily explained by assigning 

 to the germ plasm a single representative 

 for each character. As the study of 

 these hybrids was continued with more 

 minuteness, and crosses were made 

 with types possessing characters dif- 



ferent from either of the original 

 alternative pair, dihybrid ratios were 

 reported with increasing frequency. A 

 dihybrid ratio is interpreted as indicat- 

 ing that the character, instead of being 

 simple, is in reality a combination of two 

 sub-characters or factors. These factors 

 when considered singly still behave as 

 simple Mendelian units, but the com- 

 bination necessary to produce the 

 original character will occur in other 

 than the 3 to 1 ratio. The instances 

 where the slightly more complicated 

 dihybrid ratios were followed with 

 approximate exactness were sufficiently 

 niunerous to warrant continuance in 

 the belief that the law of complete and 

 equal segregation was of very general 

 application. But there were also many 

 alternative characters which showed by 

 the still more complicated hybrid ratios 

 that the original characters must be 

 represented by three, four or even more 

 factors. With more than three factors, 

 it has seldom been possible to conduct 

 experiments on a sufficiently large scale 

 to determine whether the ratios are 

 followed with exactness or not. The 

 fact that individuals occur in the later 

 generations of the cross which reproduce 

 the character in its original form, and 

 that some of these breed true, at least 

 for several generations, affords evidence 

 that we are still dealing with a com- 

 plicated form of the originally simple 

 law. 



THE LIMIT OF ANALYSIS. 



From characters composed of only 

 three or four sub-characters, there is 

 now perfect gradation to characters, 

 such as height and weight, where the 

 different stages, if they exi.st, are com- 



425 



