October 14, 1920] 



NATURE 



225 



one and the sanu jactor, anuly:>is in the stoclc shaws 

 that the glabrous condition results if any factor out 

 of a group of four is represented by its recessive 

 allelomorph. Hence we describe hairiness in the 

 latter case as a four-factor character. 



It will be apparent from the cases cited that we 

 cannot infer from the genetic analysis of one type 

 that the factorial relations involved are the same for 

 the corresponding character in another. That this 

 should be so in wholly unrelated plants is not, 

 perhaps, surprising, but we find it to be true also 

 where the nature of the characteristic and the rela- 

 tionship of the types might have led us to expect 

 uniformity. This is well seen in the case of a 

 morphological feature distinctive) of the N.O. 

 Gramineae. The leaf is normally ligulate, but indi- 

 viduals are occasionally met with in which the ligule 

 is wanting. In these plants, as a consequence, the 

 leaf-blade stands nearly erect instead of spreading 

 out horizontally. Nilsson-Ehle ( " Kreuzungsunter- 

 suchungen an Hafer und Weizen," Lund, 1909) dis- 

 covered that in oats there are at least four, and 

 possibly five, distinct factors determining ligule forma- 

 tion, all with equal potentialities in this direction. 

 Hence, only when the complete series is lacking is 

 the ligule wanting. In mixed families the proportion 

 of ligulate to non-ligulate individuals depends upon 

 the number of these ligule-producing factors con- 

 tained in the domin.int parent. Emerson (.Annual 

 Report of the Agricultural Experiment Station of the 

 University of Nebraska, 1912) found, on the other 

 hand, that in maize mixed families showed constantly 

 a 3 10 I ratio, indicating the existence of only one 

 controlling factor. 



From time to time the objection has been raised 

 that the Mendelian tvpe of inheritance is not ex- 

 hibited in the case o( specific characters. That no 

 such sharp line of distinction can be drawn between 

 the behaviour of varietal and specific features has 

 been repeatedly demonstrated. .As a case in point, 

 and one of the earliest in which clear proof of Men- 

 delian segregation was obtained, we may instance 

 Datura. The two forms, D. stramonium and D. 

 latula, are ranked by all systematists as distinct 

 species. .Among other specific differences is the 

 flower colour. The one form has purple flowers, the 

 other pure white. In the case of both species a 

 variety inermis is known in which the prickles charac- 

 teristic of the fruit in the type are wanting. It has 

 b«.'en found that in whatever way the two pairs of 

 opposite characters are combined in a cross between 

 the s[)eries, the F, generation is mixed, comprising 

 the fou4rt)ossible combinations in the proportions 

 which we should expect in the case of two inde- 

 pendently inherited pairs of characters, when each 

 pair of opposites shows the dominant-recessive rela- 

 tion. Segregation is as sharp and clean in the 

 sp«:cific character flower colour a-, in the varietal 

 character of the fruit. 



.Among the latest addition- '. li-i of 



■specific hybrids showing Mendelian inheritance, 

 those occurring in the genus Salix are of 

 special interest, since heretofore the data available 

 liiid been interpreted as conflicting with the Men- 

 delian conception. The re<ent observations of Heri- 

 bert-Nilsson (" Experimentelle Studien iiber Varla- 

 bilitjit, Spaltung, .Artbildung und Evolution in dor 

 Oattung Salix," 1018) show that those characters 

 which are regarded by svstematists as constituting 

 the most distinctive marks of the species arc refer- 

 able to an extremely simple factorial svsfem, and 

 that the factors mendelise In the ordinary wav. 

 Furthermore, these specific-character factors control 

 not only the lar^ constant morphological features, but 



NO. 2659, VOL. 106] 



also fundamental reactions such as those determining 

 the condition of physiological equilibrium and vitality 

 in general. In so far as any distinction can be 

 drawn between the behaviour of factors determining 

 the varietal as opposed to the specific characters ot 

 the systematist, Het'ibert-Nilsson concludes that the 

 former are more localised in their action, while the 

 latter produce more diffuse results, which may affect, 

 almost all the organs and functions of the individual, 

 and thus bring about striking alterations in the 

 general appearance. S. caprea, for example, is 

 regarded as the reaction product of two distinct 

 factors which together control the leaf-breadth 

 character, and also affect, each separately and 

 in a different way, leaf form, leaf colour, height, and 

 the periodicity of certain phases. We cannot, how- 

 ever, draw a hard-and-fast line between the two 

 categories. The factor controlling a varietal charac- 

 teristic often produces effects in different parts of the 

 plant. For example, the factors which lead to the 

 prculuction of a coloured flower no doubt also in 

 certain cases cause the tinging seen in the vegetative 

 organs, and affect the colour of the seed. Heribert- 

 Nilsson suggests that fertilitv between species is a 

 m.itter of close similarity in genotypic (factorial) con- 

 stitution rather than of outward morphological 

 resemblance. Forms sundered by the systematist on 

 the ground of gross differences in certain anatomical 

 features may prove to be more akin, more compatible 

 in constitution, than others held to be more nearly 

 related because the differentiating factors happen to 

 control less conspicuous features. 



Turning to the consideration of the more complex 

 types of inheritance already referred to, we find 

 numerous instances where a somatic character shows 

 a certain degree of coupling or linkage with another 

 perhaps wholly unrelated character. This pheno- 

 menon becomes still further complicated when, as is 

 now known to occur fairly frequently, somatic 

 characters are linked also with the sex character. 

 Th<' results of such linkages appear in the altered 

 proportions in which the various combinations of the 

 several characters app<>ar on cross-brwding. Linkage 

 of somatic characters can be readily demonstrated in 

 the garden stock. Some strains have flowers with 

 deeply coloured sap, e.g. full red or purple ; others 

 are of a pale shade, such as a light purple or flesh 

 colour. In most comtnercial strains the "eve" of 

 the flower is white owing to absence of colour in the 

 plaslids, but in some the plastids are cream-coloured, 

 causing the sap colour to appear of a much richer 

 hue and giving a cream "eye." Cream plastid 

 colour is recessive to while, and the de«^p sap colours 

 are recessive to the pale. When a creain-eyed strain 

 lacking the pale factor is bred with a white-eved 

 plant of some pale shade, the four possible combina- 

 tions appear in F„ but not, as we should expect in 

 the case of two independently inherited one-factor 

 characters, in the proportions 9:3:3:1, with the 

 double recessive as the least abundant of the four 

 forms. We find instead that the double dominant and 

 the double recessive are both in excess of rxiiectafion, 

 the latter being more abundant than either of the 

 combinations of one dominant character with one 

 recessive. The two forms which preponderate are 

 those which exhibit the combinations seen in the 

 parents, the two smaller categories are those repre- 

 sentinc the new combinations of one paternal with 

 one maternal characteristic. In the swe«'t pea several 

 characters are linked in this manner, viz. flower 

 colour with pollen shape, flower colour with form of 

 standard, pollen shape with form of standard, colour 

 of leaf axil with functioning caparitv of the anthers. 

 If in these cases the cross happens to be made In 

 such a way that the two dominant characters arc 



