The peculiar complication of the chemical complex determining sex in these mammalian forms, involving all the glands and hence the entire body, makes it problematical whether a complete (functional) reversal is possible, at least after any development whatever of the embryo has taken place.  On the other hand, the fact that such complete transformations have not so far been observed by no means proves their non-existence.  Their being functional, and hence to all external appearances normal, would cause such animals to escape observation.

Latent traits of the opposite sex of course immediately suggest recessive or unexpressed characters in the well-known Mendelian inheritance phenomena.  In the bird-castration cases, we saw that to remove the inhibiting sex glands caused previously latent characters to act like dominant or expressed ones.  The case of horns in sheep, investigated by Professor Wood[16], is so similar that it seems worth summarizing, by way of illustration.

Both sexes in Dorset sheep have well-developed horns; in the Suffolk breed both sexes are hornless.  If the breeds are crossed, all the rams in the first (hybrid) generation have horns and all the ewes are hornless.  If these hybrids are mated, the resulting male offspring averages three horned to one hornless; but the females are the reverse of this ratio—­one horned to three hornless.  This is an example of Mendel’s principle of segregation—­factors may be mixed in breeding, but they do not lose their identity, and hence tend to be sorted out or segregated again in succeeding generations.

In the horned Dorsets, we must suppose that both males and females carry a dual factor for horns—­technically, are homozygous for horns.  The hornless Suffolks, on the contrary, are homozygous for absence of horns.  Thus the dual factor in the zygotes or fertilized eggs at the basis of the first filial (hybrid) generation consists of a single factor for horns and a single factor for their absence.  If we represent horns by H and absence of horns by A, Dorsets have a factor HH, Suffolks AA and the hybrids HA.

All the males in this generation have horns, which means that a single “dose” of the factor H will produce horns in a male, or that they are dominant in males.  But a single dose will not produce horns in a female—­that is, horns are recessive in females—­the factor is present but unexpressed.

Mating two HA hybrids, the H and A of course split apart in the formation of the gametes, as the HH and AA did in the previous generation; so that we get an equal number of single H and A factors.  In reuniting in fertilized eggs, the chance is just half and half that an H will unite with another H or with an A—­that an A will unite with an H or another A.  Thus we have two chances of getting HA to each chance of getting either AA or HH.  Half the zygotes will be HA, one-fourth HH and one-fourth AA.