If the favoured variety is dominant it must be borne in mind that it can be either homozygous or heterozygous—that for these purposes it is represented in the left-hand column by the hybrids as well as by the homozygous dominant. In a population in equilibrium which contains about 2% of a dominant form, the great bulk of these dominants will be heterozygous, and the relative proportion of recessives, heterozygous, and homozygous dominants is given in the second line of the left-hand column.

Let us suppose now that we want to know what will be the percentage of dominants after 1000 generations if they form 2% of the population to start with, and if, during this period, they have been favoured with a 1% selection advantage. After 165 generations the proportion of recessives is 90.7, so that the proportion of dominants has risen to over 9%; after 153 further generations the percentage of dominants becomes 27.7 + 2.8 = 30.5; after 739 generations it is 88.8%, and after 1122 generations it is 69.0 + 27.7 = 96.7. Hence the answer to our question will be between 89% and 97%, but nearer to the latter figure than the former.

Mr Norton has informed me that the figures in the table are accurate to within about 5%.

APPENDIX II

The genus Papilio is a large and heterogeneous collection. It was pointed out by Haase[[93]] that it falls into three distinct sections, of which one—the Pharmacophagus section—provides those members which serve as models in mimicry; while in the other two sections are found mimics, either of Pharmacophagus Swallow-tails, or of models belonging to other groups. Though Haase's terms have not yet come into general use with systematists, there is little doubt that the genus Papilio as it now stands must eventually be broken up on these lines. To say that one species of Papilio mimics another is therefore somewhat misleading; for the differences between the Pharmacophagus group and the other two are such as to constitute at any rate generic distinction in other groups. For convenience of reference a table has been added in which the various Papilios mentioned in the text have been assigned to their appropriate sections, and referred to their respective models.

Pharmacophagus Papilio Cosmodesmus
(POISON-EATERS) (FLUTED SWALLOW-TAILS) (KITE SWALLOW-TAILS)
Antennae without scales. Antennae without scales. Antennae scaled on upper side.
Outer ventral row of spines of tarsi not separated from the dorsal spines by a spineless longitudinal depression. Outer ventral row of spines of tarsi separated from the dorsal spines by a spineless longitudinal depression. As in Papilio.
Larva covered with short hairs—with fleshy tubercles but no spines. Larva either smooth or with hard spiny tubercles. Third and fourth thoracic segments enlarged. Larva with third thoracic segment enlarged (known only in a few species).
Pupa with row of well-marked humps on each side of abdomen. Pupa wrinkled—generally with short dorsal horn. Humps if present very short. Pupa short with long four-sided thoracic horn.
Larva feeds on Aristolochia. Larva does not feed on Aristolochia. As in Papilio.
Abdominal margin of hind wing curved downwards forming a kind of groove. No scent organ. Abdominal margin of hind wing bent over in ♂ and with scent organ in fold so formed.