DISCOVERY 



329 



will take part in the formation of seed. Simple 

 mathematics enable the following scheme to be drawn 

 up: 



A Cross between Two Parents which differ in Two Pairs of 

 Characters. 



Fo . 



Tall \ ^ /Short 



Round seedJ IWrinkled seed 



Here the ratio of forms in F^ is 9 : 3 ; 3 : i and is alwa\'s 

 the same for this cross. For T and R independently 

 the ratio is, as before, 1:2:1, e.g. forms with RR ; 

 forms with Rr : forms with rr = 4:8:4 = i:2: i. 



The forms marked * are the only ones homozygous 

 in both pairs of characters, and therefore they alone 

 w^ill breed true to both. Of the remainder some, e.g. 

 TtRR, are homozygous for one pair and one. TtRr, for 

 neither. For crop production only true-breeding forms 

 are acceptable, i.e. forms homozygous in all character 

 pairs, for they alone will remain unchanged. The 

 forms of ttRR and TTrr are completely homozygous 

 and are different from both the parents (Fq). They 

 are thus examples of new forms from old ones. Where 

 the Fq differ by 3, 4, . . . pairs of characters precisely 

 the same principles apply, and it will now be realised 

 how, by choosing suitable parents, we can theoretically 

 synthesise any desired combination of characters. 



There are two great limitations to this alluring 

 theoretical possibility'. Plants will not intercross at 

 random. Generally, different forms of the same 

 species present no difficulty. But we often desire to 

 combine the characters of two different species. The 

 attempt has many times been made. In some cases 

 no hybrid seed is formed ; in others an Fj plant is 

 obtained but, like the mule, is sterile ; in yet others 

 the Fj produces some seed, but most of it fails or 

 develops into sterile F2 plants. It seems improbable 

 that new and improved crop-plants will be obtained 

 from inter-species crossing, but many forms of violas, 

 saxifrages, roses, irises, and pinks have been obtained 



in this way, w-hile among the orchids even inter-generic 

 crosses have succeeded. Strange to say, wheat and 

 rye are partially inter-fertile. 



" Linkage " is a phenomenon which imposes the 

 second limitation to the welding together at will of 

 plant characters. This in effect is a tendency of two 

 pairs of characters to cling together, so that in the F^ 

 of a cross there is not the uniform assortment of 

 character combinations which was described above for 

 unlinked characters. Certain combinations maj^ appear 

 in F2 only in the proportion of, say, 1/1,000 or even not 

 at all. Linkage is an e.xception to Mendel's second 

 law which he himself never encountered. 



Mendelism and the Farmer 



The British farmer grows many plant forms of 

 hybrid origin. He does not always realise this, and 

 when confronted with an admitted product of 

 methodical hybridisation often asks, " Will it hark 

 back ? " — i.e. will it in time revert to the form of its 

 less valuable parents ? If the work has been weU done, 

 if the new form is homozygous in all characters, the 

 answer is an emphatic No. Such a form, if protected 

 from foreign pollen, will continue to breed true. Plant 

 forms, whether natural or derived from artificial 

 hybridisation, homozygous in all characters to start 

 with, and safeguarded from natural cross-pollination, 

 are called " Pure Lines." Being immutable, they 

 cannot be improved by selection. For a hundred years 

 past, and even to this day, time and effort have been 

 vainly spent in attempts to improve pure lines by 

 selection. The most common fallacy has been that 

 the continued selection year after year of the largest 

 seeds must " improve the type." But careful experi- 

 ment has shown that if, year after year, the largest 

 and the smallest seeds of the crop of a pure line are 

 separately selected and grown, the resultant plants 

 are the same. Commercial stocks of seed often contain 

 five or six different pure lines, and selection will 

 separate these out but can do no more. Catalogues 

 often expatiate on " Jones's re-selected variety." At 

 the best this can only be a stock free from admixture 

 with other forms. Such a stock is valuable, but not 

 always to the extent of the enhanced price demanded. 



The immutability of the pure line has one exception. 

 Among the most carefully safeguarded stocks may be 

 one or two plants different from the type. They are 

 called " mutations," and arise from the original form 

 by some as yet inexplicable process of nature. Mutant 

 buds occur on some trees and may be propagated by 

 budding on to a suitable stock. Mutations are 

 exceedingly rare, and accidental admixture of seed or 

 natural cross-pollination is not infrequently the cause 

 of their reported occurrence. 



In many minds the comparative simplicity of the 



