Cases in which the first stage is variable appear to be of the least frequent occurrence. I know of only one such instance, viz., _Anceryx Pinastri_, of which the newly hatched larvæ (Pl. VI., Fig. 53) show considerable differences in the brownish-black crescentic spots. The second (Fig. 54), third, and fourth stages are then tolerably constant, while the fifth stage again is very variable. An instance of this kind can be easily explained by two waves of variation, the first of which now affects only the first stage, while the second has just commenced to affect the fifth stage. Such a supposition is not opposed to any theoretical considerations, but rather has much probability in its favour, since we know that species are from time to time subject to be remodelled; and further, that the coalescence of several stages of phyletic development in the ontogeny of one and the same species (see p. 226, development of the genus _Deilephila_) shows that during the backward transference of one character, new characters may appear in the last stage of the ontogeny, and indeed very frequently at a time when the next youngest character has not been transferred back so far as to the first stage. That this secondary variability is to a certain extent brought about by the conflict between the old and new characters, the latter striving to suppress the former, is shown by the caterpillar of _Saturnia Carpini_ which I have observed for many years from this point of view, and than which I do not know a more beautiful illustration. When these larvæ leave the egg they are black, but in the adult state are almost bright green--this at least being the case in a local form which, from the district in the vicinity of Genoa where it is found, I will designate as the var. _Ligurica_. Now whilst these two extreme stages of development are relatively constant, the intermediate stages show a variability which becomes greater the nearer the last stage is approached, this variation in the marking depending simply on the struggle between the green colour and the more anciently inherited black. In this manner there arises, especially in the fourth stage of the German local form, an incredible mixture of the most diverse markings, all of which can, however, be very easily explained from the foregoing point of view. The simpler and, as I am inclined to believe, the older form of the transformation is presented to us in the local variety _Ligurica_. In the last stage, when 7.5 centimeters long, this form is of a beautiful bright green colour without any trace of black marking[174] (Pl. VIII., Fig. 77). The colour of the six orange warts which are situated on each segment is also similar in all specimens, so that this stage is perfectly constant.