In _A. Levana_ the _Levana_ form is decidedly more constant than the _Prorsa_ form. The first is, to a slight extent, sexually dimorphic, the female being light and the male dark-coloured. If we take into consideration this difference between the sexes, which also occurs to a still smaller extent in the _Prorsa_ form, the foregoing statement will be found correct, viz. that the _Levana_ form varies but little, and in all cases considerably less than the _Prorsa_ form, in which the greatest differences occur in the yellow stripes and in the disappearance of the black spots on the white band of the hind wing, these black spots being persistent _Levana_ markings. It is, in fact, difficult to find two perfectly similar individuals of the _Prorsa_ form. It must, moreover, be considered that the _Levana_ marking, being the more complicated, would the more readily show variation. Precisely the same thing occurs in _Pieris Napi_, in which also the var. _Æstiva_ is considerably more variable than the var. _Vernalis_. From the behaviour of the var. _Bryoniæ_, on the other hand, which I regard as the parent-form, one might be tempted to raise an objection to the theory; for this form is well known to be extraordinarily variable in colour and marking, both in the Alps and Jura, where it is met with at the greatest altitudes. According to the theory, _Bryoniæ_ should be less variable than the winter form of the lowlands, because it is the older, and should therefore be the more constant in its characters. It must not be forgotten, however, that the variability of a species may not only originate in the one familiar manner of unequal response of the individual to the action of varying exciting causes, but also by the crossing of two varieties separately established in adjacent districts and subsequently brought into contact. In the Alps and Jura the ordinary form of _Napi_ swarms everywhere from the plains towards the habitats of _Bryoniæ_, so that a crossing of the two forms may occasionally, or even frequently, take place; and it is not astonishing if in some places (Meiringen, for example) a perfect series of intermediate forms between _Napi_ and _Bryoniæ_ is met with. That crossing is the cause of the great variability of _Bryoniæ_ in the Alpine districts, is proved by the fact that in the Polar regions this form “is by no means so variable as in the Alps, but, judging from about forty to fifty Norwegian specimens, is rather constant.” My friend, Dr. Staudinger, who has twice spent the summer in Lapland, thus writes in reply to my question. A crossing with _Napi_ cannot there take place, as this form is never met with, so that the ancient parent-form _Bryoniæ_ has been able to preserve its original constancy. In this case also the facts thus accord with the requirements of the theory.