THE CO-OPERATION AND NATURE OF GENES Epistasy.—In a normal organism very many genes must co operate to produce the normal structures, chemical substances and functions. Thus in the rabbit at least io genes must be pres ent to produce the normal coat colour. The most important of these is C. A rabbit of the ultimate recessive genotype cc of the series of allelomorphs to which it belongs is a pink-eyed white, and all such rabbits are indistinguishable, no matter what other genes determining colour they may carry. Thus if a white rabbit carries two of the genes 1 for density of pigment, it will give dense-coloured young when mated with a pure-bred blue (CCii) ; if it is recessive it will give blue on some other dilute colour, if heterozygous, half dense and half dilute. But 11 and ii whites can only be distinguished by their ancestry or descendants. A gene such as 1 which can only manifest itself in presence of an other such as C is said to be epistatic to it. All other colour genes are epistatic to C.
A rabbit of the composition eegg, but bearing the other colour genes of the wild rabbit, is born yellow but de velops some black pigment, becoming a "tortoiseshell." EEgg or Eegg is black, E causing extension of the dark eye-pigment into the coat; eeGG or eeGg is a pure yellow, G inhibiting the black pigment in the coat ; EEGG, EeGG, EEGg or EeGg is the wild col oured grey. Here G inhibits all pigmentation on the belly, and also black pigmentation on portions of the hairs, which have there fore yellow tips. Such colour inhibitors are very common. Thus several types of dominant white fowl are known, in which pig mentation has been inhibited by different genes. Similarly there is a dominant white form of Primula sinensis, and Primula elatior contains a gene which, on crossing the species, inhibits the an thocyanin colour of Primula Juliae, although P. elatior itself does not contain any genes causing the formation of anthocyanin against which the inhibitory gene could act. Again Drosophila melanogaster (and doubtless other Diptera) contains two inde pendent genes inhibiting the growth of the balancers or rudi mentary wings. In the absence of either, these develop into wing like structures.
Several genes may co-operate to produce a given effect. Thus in the sweet pea there are two types of recessive whites, due to genes in different chromosomes. Each breeds true but when crossed they give coloured flowers. Calling the colour genes C and R the heterozygous coloured form is CcRr. When selfed it gives an
of 9 CR : 3Crr : 3ccR : iccrr, i.e., 9 coloured to 7 white. Again any of a number of genes may be sufficient to produce a given effect. Thus in hexaploid wheats any of three genes will produce red colour. If a plant containing all three be crossed with a triple recessive, only i in 64 of the
will be white. Still more compli cated cases are known, and they constitute one of the greatest practical difficulties in the study of heredity. The case in which the effects of the several genes are not clear-cut will be considered later.
Many genes, though of course inherited through both sexes, can only find expression in one. Thus milk-production and butter-percentage in cows, and egg laying in hens, are both determined by genes inherited from the father as well as from the mother. The same is true of characters not obviously connected with sex. For example in Drosophila a particular recessive gene, "bobbed," causes a shortening of bristles in the female only. This is because the Y-chromosome normally contains the dominant allelomorph of "bobbed." When this is lost the character appears in both sexes. A gene which only acts upon one sex is described as sex-limited in its expression. Such genes are also sometimes, though not generally, sex-linked. The action of a gene may be delayed for a generation. Thus a particular gene in the sweet pea causes all the pollen grains produced by plants carrying it to be long, as opposed to round. As this gene is domi nant, the pollen grains of a heterozygous individual are all long, even though half of them carry the recessive gene for round pollen. Similarly a gene in the silk-worm causes females bearing it to produce pink eggs and larvae, so that here also the appearance of the larvae is no guide to their genetic composition. Such inheri tance is called maternal, and is a special (and in practice rather confusing) type of sex-limited inheritance.
as genes are influenced by one another in their expression, they are influenced by variations in the environment. For example the size or flowering time of an organism are naturally determined by its nutrition and other envi ronmental factors, as well as by genes. But very often the effect of environment is much greater on one phenotype than on another. Thus cold increases the amount of black pigment in the skin and hair of the "Himalayan" rabbit, where the blackness of the extrem ities is only due to their low temperature ; but it has little or no effect either on the white or the fully coloured rabbit. Again whereas damp food has very little effect on normal Drosophila, it causes the production of grossly abnormal abdomens in a particu lar genotype. Where it is desired to eliminate a hereditary char acter, the environment should be so adjusted that it can find its fullest expression. Thus the seed of many beets and marigolds, when sown in the open, produces about 1% of bolters, i.e., plants flowering in their first year. Bolting depends both on heredity and environment. If the seed is sown under glass in December and planted out in the spring, half or more of the plants may bolt, and by eliminating them for several generations the hereditary ten dency to bolt may be got rid of, which is almost impossible under ordinary conditions of planting.
The simplest explanation of the phenomenon of dominance is the "presence and absence" theory. According to this view a recessive gene is simply the absence of the corresponding dominant gene. One need not of course imagine a mere void in its place, but some structure which does not have its function. Where dominance is complete it is assumed that one of the dominant genes can effectively do the work of two. In a series of multiple allelomorphs it is supposed that the various intermediate members of the series function to a lesser degree than the ultimate dominant. In this case we should expect that the intermediate genes would not be completely domi nant. This is generally the case. Thus in Drosophila melanogaster there are 10 genes in the same locus giving eye colours intermediate between white and the normal red. None of these show the com plete dominance which red shows over then all. If an activity of the gene or genes arbitrarily denoted by r oo or over is needed for complete colour, it is clear that no gene producing by itself an effect of less than ioo could be fully dominant. The presence and absence theory is borne out by the facts of sex linkage and deficiency, where an absent gene behaves like a recessive gene. But it meets with difficulties in explaining the origin of genes dominant over the normal type, and occasionally breaks down where, in a triploid, two recessive genes are dominant over one which is dominant in the diploid. Nevertheless, in a modified form, it is very widely applicable.
In some cases we may say that a dominant gene causes the formation of a certain substance, e.g., a particular anthocyanin pigment or a particular colloid concerned in immunity. Thus in the guinea-pig the presence of complement (see IMMUNITY) in the serum is dominant to its absence. The recessives are so liable to disease that the character has died out. More rarely a gene inhibits such a formation. On the presence and absence theory it should do so by producing a definite substance. This is at least sometimes the case. Thus the skins of black rabbits contain a cata lyst which causes the formation of melanin (black pigment) from tyrosin and hydrogen peroxide. This is absent in recessive white and yellow rabbits. The gene G which inhibits pigment formation produces a substance in the skin of the white bellies of rabbits carrying it which can be shown in the test-tube to inhibit pigment formation by the extract of black skin. A dominant gene causing the "English" piebald pattern acts by forming the same or a simi lar inhibitory substance. The co-operation of genes is therefore largely to be explained on biochemical lines.
In some cases the genotype determines the rate of a reaction, such as the growth of a given organ, or the rate of formation of a pigment. Thus a series of genotypes in certain insects and crustaceans are characterized by different rates of darkening during larval life. Those which arrive at a certain stage before the end of metamorphosis are indistinguishable in the adult form. But those characterized by very slow rates cannot complete pigment formation before development is over, and yield light coloured adults.
The gene then is a unit, situated (dur ing nuclear division) at a definite point in a definite chromosome, and dividing once at each cell division. Its diameter, to judge from data as to linkage, is probably not much more than ten times that of an average protein molecule, and may be less. A gene may be responsible for any of the differences which exist between different varieties of a species, including types so aberrant as to be incapable of life. There is reason to suppose that genes act by promoting or retarding specific chemical reactions. The heads of spermatozoa consist largely of compounds of nucleic acid with rather simple proteins.
