September 30, 1922] 



NA TURE 



46; 



The Inheritance of Size. 1 



"THE stud}- of size-inheritance is beset with diffi- 

 -*■ cutties which do not attend the study of 

 qualitative differences involving colour and form. 

 Students of genetics have stated size-inheritance in 

 various plants and animals in terms of multiple 

 size factors segregating independently in the Mendclian 

 fashion ; but the universal presence of fluctuations 

 which obscure the quantitative effects of separate 

 factors, as well as other difficulties, have prevented 

 the studv of size factors being in the same satis- 

 factory condition as that of the factors which control 

 sharply marked qualitative characters. 



Stature in man has been investigated from the 

 time of Quetelet and Gabon to the recent paper of 

 Davenport, 2 but investigators are not yet agreed 

 even concerning the nature of the Mendelian units, 

 if such they be, which affect and control this feature 

 of bodily measurement. Are there only general 

 growth factors, or are there also separate factors 

 influencing the length of individual segments of 

 the body, such as the legs, trunk, and neck ? Daven- 

 port concludes that both types of factors are present, 

 and that some races and families have different 

 relative lengths of these segments because of the 

 independent inheritance of such local factors con- 

 trolling the length of individual bones or segments. 

 Moreover, Davenport believes that crossing between 

 races leads to various bodily disharmonies, such 

 as large teeth in small jaws or a small heart in a 

 large body. 



Castle, in a recent study of size-inheritance in 

 rabbits, 3 criticises Davenport's view of local size 

 factors as essentially preformationist, and shows 

 with considerable success that, so far as the rabbits 

 of his breeding experiments are concerned, general 

 inherited growth factors appear to control the size 

 reached by all parts. In crosses between the large 

 Flemish rabbit and small varieties such as the Polish 

 and Himalayan, Castle concludes, as Punnett and 

 Bailey 4 had concluded from earlier experiments on 

 weight in rabbits, that several size factors are in- 

 volved, as indicated by the greater range of variation 

 in F, and later generations than in F\. This sub- 

 stantiates other results of these authors 5 with 

 poultry. They crossed Gold - pencilled Hamburghs 

 and Silver Seabright Bantams and obtained in F 2 

 and F 3 both larger and smaller birds than the original 

 parental types. That several independent factors 

 are concerned in the determination of size or weight 

 in birds and mammals seems then well established. 



But another difficulty comes in to obscure such 

 quantitative results, and that is the fact of hybrid 

 vigour or heterosis, which occurs largely or entirely 

 in the F a generation, producing a general increase in 

 the size of the F, offspring. For example, in the 

 rabbit crosses, the F x is nearer the size of the larger 

 parent owing to this effect, but the effect disappears 

 in the F 2 and later generations. This of course 

 shifts the curve of size temporarily towards the 

 right. 



Castle made a careful study of the growth-curves 

 of his rabbits, weighing them at intervals throughout 

 their development to maturity, but he appears not 

 to have studied the variation of his races before 



1 " Genetic Studies of Rabbits and Rats." By W. E. Castle. (Publication 

 No. 320.) Pp. 55. (Washington: Carnegie 'Institution, 1922.) 1 dollar. 



2 Davenport, C. B., 1917, " Inheritance of Stature," Genetics, 2, pp. 313- 

 389- 



3 Castle, \V. E.. 1922, " Geneti- Studies of Rabbits," etc. 



* Punnett, R. C., and Bailey. P. G., 101S, '-Genetic Studies in Rabbits: 

 I. On the Inheritance ot Weight/' Journ. Genetics, 8, 1-25. 



5 Punnett, R. C, and Bailey, P. G., 1914, "On Inheritance of Weight in 

 Poultry," Journ. Genetics, 4, pp. 23-39. 



crossing. He concludes that the adult Flemish rabbit 

 is larger because it is larger at birth and grows more 

 rapidly and for a longer period than the small Polish 

 rabbit. This is contrary to the views of Punnett 

 and Bailey that age of maturity is not necessarily 

 closely correlated with size. Castle applies his results 

 to man, and reasons that natives of the south of Italy 

 are short of stature and short-limbed because they 

 cease to grow at a relatively early age, while 

 Swedes and Scotch are tall and long-limbed because 

 they mature later. 



In the hybrid rabbits, series of measurements 

 were made of weight, ear-length, skull dimensions, 

 and certain leg bones. From these data the correla- 

 tion-coefficients between the various measurements 

 were determined and were found to be uniformly 

 high. Thus the correlation between ear-length and 

 weight in F, and F 2 rabbits was 0-836, and between 

 lengths of femur and skull 0-871. This furnishes 

 strong support for the conclusion that the size of 

 all parts is determined by general growth factors 

 affecting the whole body, and not by independently 

 segregating factors affecting the size of particular 

 organs. Davenport points out that certain races 

 of man have long legs and relatively short trunks, 

 while others have short legs and longer trunks, but 

 Castle holds that the former races are absolutely 

 taller, and regards them as a later growth stage than 

 the short races. Whether this explanation will 

 apply to all races of man remains to be seen. The 

 most urgent requirement at the present time is a 

 mass of accurate anthropometric measurements of 

 all parts of the body in various races. 



Many genetic factors are known to affect chiefly 

 one organ of the body, such as the eye or the wing in 

 flies, and since that is the case there seems no a priori 

 reason why some size factors should not also affect 

 chiefly certain organs. To demonstrate such an 

 effect, however, a considerable mass of biometric 

 data is required. So far as plants are concerned, 

 the results of Gates 6 show that size factors in hybrids 

 are in some cases local in their effects. In crossing 

 species of CEnothera having large and small flowers 

 respectively, he obtained in F 2 and later generations 

 frequently a wide range of flower-size on the same 

 plant, and in many cases even the four petals of the 

 same flower differed widely in length. Thus it is 

 clear that local size factors occur in plants. Whether 

 they also occur in animals and man remains to be 

 determined. 



That an increase in the range of variation of the 

 F 2 as compared with the F x is not in itself sufficient 

 to" prove the presence of several inherited size-factors, 

 is indicated by a recent paper of Sumner and Huestis. 7 

 In connexion with extensive breeding investigations 

 of the California deer-mouse, Peromyscus maniculatus, 

 they have compared the length or weight of corre- 

 sponding right and left bones such as the mandible 

 and femur. In this way they obtained sinistro- 

 dextral ratios for these bones and showed statistically 

 that there is no inheritance of such a ratio from one 

 generation to the next, e.g. if the parents had a 

 slightlv longer left femur there is no tendency for the 

 same 'condition to be repeated in the offspring. 

 Nevertheless, they found that in crosses between 

 different sub-species these ratios showed greater 

 variability in F 2 than in F 1 . This fact will need to be 

 taken into account in future studies of size- variation. 



■ Gates, R. R., r.917, "Vegetative Segregation in a Hybrid Race," Journ. 



f. pp. 237-253. 

 ' Sumner, F. B., and Huestis, R. R., 1921, " Bilateral Symmetry m its 

 1 to certain Problems of Genetics," Genetics, 6, pp. 445-4S;. 



NO. 2761, VOL. I IO] 



