﻿NORTH AMERICAN SPHINGID^E. 101 



we have no special investigations to offer on the question, we can at least clear and 

 define our conceptions with respect to what should be regarded a variety of a species. 



From what has been heretofore said respecting species, we cannot suppose a variety 

 can be detected, simply from the study of the perfect insect in its most aberrant 

 condition. For wherein does a variety differ from the species ? by what means is 

 the systematist enabled to ascertain the specific identity of those individuals, which 

 have been described as distinct species? Is it by any difference in the ovum, any 

 peculiarity in the form, structure, ornamentation or biography of the embryo, any 

 differences in pupation, or any essential or specific variation in the structure of the 

 perfect insect ? By no means. To be a variety, or wandering from a certain specific 

 type, it must observe the same biographical and organic cycle, possess the same specific 

 characteristics of structure in its perfect state, but differ from the species in its pecu- 

 liarities of ornamentation, and in its size, perhaps, to a degree that without a know- 

 ledge of its embryology and biography, it would be pronounced and registered distinct 

 from the perfect individuals towards which it shows the strongest specific affinities 

 of structure. Another characteristic of the variety is, that there is no stability even 

 in the peculiarities of its ornamentation ; and whether it occurs in the same brood 

 under identical climatal conditions, or is found as an isolated perfect being in a widely 

 separated geographical area under dissimilar conditions, it must be associated with its 

 normal type ; and that the ova of the variety will reproduce, not only this, but also 

 more or less aberrant perfect individuals. An invariable, fixed and constantly re- 

 curring ornamentation in any group of perfect beings, except when it is a mere intensity 

 or pallidness of hue, which will but rarely mislead, is not only incompatible with 

 the conception of a variety, but would constitute a true species even when their biog- 

 raphies are closely coincident. Variation, or specific instability, observing fixed and 

 determined limits which must be ascertained by observation, is part of the true 

 history of species. It is not manifested to the same degree, probably, in the specific life 

 of every true species, but wheresoever and whensoever it does occur, is capable of 

 being referred to its normal type, by its agreement in all those essential characteristics 

 necessary to form a conception of true species. As long as the specific diagnosis 

 must be confined to a description of the ornamentation of the perfect being, there 

 are no means of distinguishing certainly the variety from species, should the former 

 differ from the latter essentially in this respect, and I have no doubt that every effort at 

 systemization with a knowledge of perfect forms alone, contains many illustrations 

 of the attendant difficulties of discrimination. 



It is not my desire to enter into any special discussion of this subject ; it is sufficient 



to enunciate what I believe truth, and apparent to my own mind. Discussions must 



be founded either on principles or prejudices; if the former should differ, there can 



2G 



