The genilalia of animals, particularly insects, are extremely useful taxonomic characters, for they are highly conserved within species yet generally diverge rapidly during specialion: often the only reliable means of separating sibling species is via the morphology of the genilalia. A number of explanations have been proposed to explain this pattern of variation, the most prominent of which is the ‘lock-and-key’ hypothesis. This hypothesis and others are tested using the meadow brown butterfly, Maniola jurtina (L.), which exhibits variation in the male genitalia. A novel technique is described which enables dissection of copulating couples. Sampling and dissection of male butterflies from 14 populations in southern England was carried out to quantify variation in the genitalia. Mating success and the strength of the male-female bond during copulation was assessed in relation to the dimensions of the male genitalia.

The most variable portions of the male genitalia are the distal and dorsal margins of the paired valves. Contrary to the lock-and-key hypothesis, and the more recent sexual selection model, evidence is presented that the most variable portions of the genitalia have no apparent function during copulation, and therefore may not be subject to direct selection. I suggest that neutrality to selection was essential for the evolution of the wide variety in genital morphology currently found in the Lepidoptera.