XVlii INTRODUCTION. 



used as the sole basis of genera seems clear, from the fact that they 

 differ very widely in co-existing species, in all other respects closely 

 allied and generically identical. As to the case of augur and haruspica, 

 the main distinguishing peculiarities of augur are all retained, as we 

 have seen, in haruspica, except the structure of the genitalia. Since 

 there can be no question that augur and haruspica are related geneti- 

 cally, that the differences, such as they are, are the result of a 

 division of the original stock, of time and environment, we must come 

 to the conclusion that a modification of the genitalic appendages is 

 sooner effected than a change in color, pattern and form, and that 

 therefore the character is morphologically of less importance, is less 

 stable than other, apparently more superficial characters. The co- 

 existing genitalic species of Nisoniades have been separated in the 

 books, but no experiments have been made as to whether the insects 

 produce always the genitalic type ; the immature stages are not known, 

 and this latter is also the case as to haruspica and rubifera, where, if 

 known, they have not been properly compared with those of augur and 

 rubi. We may consider the whole series of genitalic studies as being 

 incomplete, as lacking confirmation and the basis of experiment, while 

 we may be willing to^admit, on the face of the matter, that they would 

 be important if true. There are two or three aspects of the case which 

 might be mentioned here. The first relates to the want of precision 

 in the forms of the male appendages, owing to which, it is difficult to 

 state what amount of modification is specific, what generic. It is clear 

 here, as elsewhere in the moths, that specific and generic characters 

 differ only in quantity, not in quality, that our categories are essen- 

 tially artificial, and do not exist in nature. The second stands still 

 clearly within a scientific or objective point of view. Some writers 

 on genitalia have endeavoured to classify the whole Lepidoptera by 

 their tails, much as Herrich-Schiiffer took only cognizance of the wings 

 and neuration. In one case, as in the other, the result must be failure. 

 The last observation I would make leaves the theoretical domain of 

 science. It refers to the overbearing language and assumption of cer- 

 tain recent authors who, on the strength of their studies as to the 

 genitalia of a few species, allow themselves to speak disparagingly of 

 the work of those who have not done the same, taking the ground, 

 apparently, that the species are alone to be determined by their geni- 

 talia, whereas this is by no means the actual state of the case. In 

 strongly differing species, the genitalia remain of the same pattern. 

 While we may admit that the genitalia are a guide in difficult cases, 

 the criterion for species remains, that the insects breed true to type ; 

 that in some, if not all stages, they may be distinguishable by experts, 

 a difference in the genitalia being only one link in the chain by which 

 we determine that we may have to do with a distinct cycle of repro- 

 ducing existences. The reality of this fact must be finally assured by 

 breeding. 



Leaving the genitalically founded representative species, inhabit- 

 ing Europe and North America, we come to those which are separable 

 chiefly by differences in the larval stage. Here we may cite the 

 European Tricena psi and the American T. occidentalis, the moths of 

 which are much alike, while the caterpillars seem to be easily dis- 

 tinguished. Here also belong a number of instances deduced by 

 Guenee from Abbot's drawings of American larvas. The immature 



