The method of inheritance of the feathering on the feet of some poultry has already been made the subject of much study. Hurst (1905, p. 152) crossed booted and non-booted birds and bred the hybrids together. He concluded that "the Mendelian principles are at work in these aberrant phenomena, but are masked by something not yet perceived." My own conclusion (1906, p. 72) was: "Booting is dominant, but usually imperfectly so." A more extended study has been desirable.

Booting is variable in amount. To indicate its degree I have had recourse to an artificial scale. I recognize 11 grades, running from 0 to 10. The grade 0 implies no feathers whatsoever. Grade 10 implies heavy booting extending over the front half of the shank. Grade 5 implies an extent of only half of the maximum, i. e., the outer front quarter of the shank. Intermediate grades indicate intermediate extension of the feathered area.

A. TYPES OF BOOTING.

The races of booted poultry used have been as follows: First, bantam Cochins of two varieties; second, a bantam Dark Brahma; and third, the Silkie. In my representatives of the first two groups, but particularly in the Dark Brahma, the amount of booting is variable. In one type the outer third of the shank in the newly hatched chick is covered by strong, heavy, specialized feathers, directed outward, while the middle and inner thirds are covered by smaller, finer, imbricating feathers sparsely placed and resembling reduced contour-feathers. In most individuals the transition from the one kind to the other is gradual, while in others it is sharp, and in a few the outer third only of the shank is feathered. In the Silkies, which the standard poultry books describe as being more sparsely feathered on the shank,[8] the outer zone of feathers is the only one developed; and, occasionally, as table 31 shows, even these feathers may be lacking. We have thus two types to distinguish—the extended (Cochin, Brahma) type and the restricted type.

B. NORMAL VARIABILITY.

An inspection of table 31 shows that, in respect to booting, the Cochins and Dark Brahmas are clearly closely related to each other. Owing to smaller numbers and to other circumstances that will be discussed later, the results are less regular in the Dark Brahma offspring, but in both the range is from 2 or 3 upward to 10, with a great preponderance in grades above 5. In the Silkies, on the other hand, the greatest frequency is found in grades below 5. This difference is correlated with a difference of the parents, for the commonest grades of the parents of the Cochins are between 6 and 10, of the Dark Brahmas between 5 and 9, and of the Silkies between 1 and 3. These results suggest that the Silkie is typically heterozygous in boot, producing 25 per cent recessives (boot of grade 4-7) and 75 per cent dominant (0, 1) and heterozygous (2, 3). We shall see that this hypothesis receives support from all Silkie matings.

Inside of any part of this table it appears that, on the whole, as the average grade of the boot in the progeny diminishes that of the parentage diminishes, although the correlation is by no means perfect. Thus the average of the parental grades in the first part of table 31, A (which is arranged in descending order of the averages of the offspring) is 8.5; in the lower half, 7.4. The average of parental grades in the upper half of table 31, B is 6.4; in the lower half 5.5. In table 31, C the grades are 2.9 and 2.3, respectively. This correlation indicates, without exactly measuring, heredity in grade of booting.

Table 32 shows the results of crosses between Cochins (high grade of boot) and Silkies (low grade).

So far as the average grade of boot in offspring goes, this table stands between that of the Cochins (table 31, A) and that of the Silkies (table 31, C). But what is especially striking is the apparent dimorphism revealed in the line of totals. There is one (empirical) mode at 10, corresponding with that of the Cochins, and a second clear mode at 5, corresponding to that of the Silkies. If we assume the Cochin to be homozygous in boot (RR) and the Silkie to be heterozygous in boot, then we can interpret the high mode as extracted recessives, the median mode as heterozygotes.

C. RESULTS OF HYBRIDIZATION.

We have next to consider the nature of the inheritance when one parent belongs to an unbooted race, the other to a booted one (table 33).

An inspection of Table 33, which gives the distribution of grades of boot in the offspring constituting the first hybrid generation, might well lead to the conclusion that inheritance is here of a blending nature, or that, if either condition is dominant, it is the booted one, as suggested in my report of 1906. On this hypothesis the offspring with no boot illustrate imperfection of dominance, and one would say that, in booting, dominance is very imperfect.

However plausible such an interpretation might appear when based on the first hybrid generation alone, it becomes untenable when subsequent generations are taken into account, as we shall see later. The hypothesis breaks down completely in the second hybrid generation and we are forced to the opposite hypothesis, namely, that the clean-shanked condition is dominant. Such an hypothesis would seem, at first, to contravene the principle enunciated in my report of 1906 that the more progressive condition is dominant over the less progressive condition, or absence. But such is not necessarily the fact. We have no right to assume that presence of boot is the new character. The rest of the body of poultry (save the head) is covered with feathers. If the foot is not it must be because there is something in the skin of the foot that inhibits the development of feathers there. And this inhibiting factor is dominant over its absence.

Table 33 shows that the Silkie crosses yield an exceptionally high per cent of the dominant clear-footed condition. This is additional evidence that the Silkies are DR, and so this cross produces 50 per cent of pure extracted dominants in addition to 50 per cent of heterozygotes in booting.

To get further light on the nature of inheritance of booting we pass to the examination of the second hybrid generation (table 34).

In the case of Silkies, which throw 67.6 per cent clean-shanked progeny in F₁, we find in F₂ only about 60 per cent clean-shanked. This diminution is, of course, due to the extraction of some pure booted recessives, which draw from the proportion of clean shanks.

In the case of the Cochins and Dark Brahmas, expectation, with perfect dominance, is that 75 per cent of the offspring shall be clean-shanked. Since dominance is imperfect (as shown by the occurrence of many booted birds in F₁) we should look for an actual failure to reach so large a proportion, but we are hardly prepared for the result that in most of the F₂ crosses of Cochins and Brahmas less than 25 per cent of the offspring are clean-shanked. In 4 pens the average is only 10 to 12 per cent, and in one only 2 per cent of the offspring fail to develop feathers on the feet. What shall we say of such a case as the last? The history of the father (No. 666) is absolutely certain; his mother was No. 121, the original Dark Brahma female, with a boot of grade 9 and a record in her immediate progeny that indicates perfect purity of booting in her germ-cells. His father was a White Leghorn with clean shanks and without a suspicion of having such antipodal blood as the Asiatic in his ancestry. No. 666 is certainly heterozygous in boot, if boot is a single unit. The hens with which No. 666 were mated were clearly heterozygous, as is known not only from their ancestry, but also from their behavior when mated with another cock, No. 254, in which case they threw 12 per cent non-booted offspring. If now both parents are heterozygous they must produce 25 per cent recessives. This is the fact that forces us to conclude that clean shank is not recessive, but dominant and due to an inhibitor that frequently fails to dominate. In table 31 the two recessive varieties, mated inter se, produce no featherless shanks; the feathers grow freely as they do over the rest of the body. Some of the Silkies of table 31, however, are really heterozygous, with the dominant inhibitor not showing; consequently they throw a large proportion of non-booted offspring. In F₁, as table 33 shows, the heterozygous offspring have a reduced boot and perfect dominance—complete inhibition of boot—in from 6 to 68 per cent. Dominance is most complete in the Silkies, where, the feathering being feeble, the inhibitor has, as it were, less to do in overcoming it. In F₂ the expected 75 per cent dominant is approached in the case of the Silkies (62 per cent and 59 per cent, respectively), but inhibition is very imperfect in the Cochin and Brahma crosses, being reduced to between 25 and 2 per cent. More proof that boot is due to the absence of a factor rather than to its presence is found in this generation. If absence of boot is recessive, then, combined with imperfection of dominance, at least 25 per cent of the offspring should be recessive and probably a much larger proportion. The results in table 34 are absolutely incompatible with this hypothesis, since, in one case, there are only 2 per cent that can not develop boot. Two extracted clean-footed birds sometimes throw boot and sometimes not, and this result is to be expected on the hypothesis that clean-footedness is dominant, but two heavily booted birds can not transmit the boot inhibitor.