546 BRITISH LEPIDOPTERA. 



culty of separation is rarely experienced in this direction. What, then, 

 must be the criteria of species here ? What keeps the species pure, 

 when they could so readily intercross ? We do not know, but we 

 imagine that the specialisation of the male abdominal scent glands 

 helps them to select their own mates, and that this is aided by the 

 isolation of the imagines by certain habits, by the specialisation of the 

 larvte to certain food-plants, and similar factors of equal importance. 

 When, therefore, a race has become sufficiently specialised by habit, 

 habitat, food-plant, or similar cause, to maintain itself as a distinct 

 form, definable structurally by some peculiarity of colour or distribu- 

 tion of markings, that race has in our opinion attained specific rank, 

 and should be so treated. That it should exhibit a marked difference in 

 egg, larva and pupa also, is hardly to be expected, except as a result of 

 long-continued isolation, and there is no reason why the genitalia of a 

 group should not be as generalised as the neuration. Specialisation 

 in other directions than the early stages antennas, neuration, geni- 

 talia, etc. must be looked upon as of the greatest importance in a group 

 like this when it does occur, but these must not be considered as the 

 only criteria in the determination of Anthrocerid species. 



ADDENDA. 



p. 869. Cochlidion avellana, Linne. In accepting avellana, Linne, 

 as the specific name for the insect more generally known as limacodes, 

 Hufn., or testudo, Fab., we followed Kirby and Werneburg. Grave 

 doubts as to whether avellana, Linne, referred to the species in 

 question, were raised when we entered into the matter more fully, and 

 we are now informed by Prout that the insect in the Linnean cabinet 

 that bears this name is a Tortricid moth. We know that the insects 

 in this cabinet are not as Linne" left them, still, in the present un- 

 certainty, we think it advisable to add Hufnagel's description. This 

 reads as follows : 



Phalaena limacodes. Die Schildmotte. Briiunlichgelb mit 2 braunen Quer- 

 streifen, die Unterfliigel etwas brauner. Raupe, glatt, kurz und dicke. Hat die 

 Figur eines Schildes ; griin mit rothen und gclben Streifen. Auf den Eichen und 

 Biichen ; September. Vogel. an den Bliittern der Biiume ; Mai und Junius. Von 

 der vierten Grosse. Selten [Berl. Mag., iii., p. 402 (1766)]. 



Hufnagel adds (loc. cit., p. 425) that when he first found the larvae 

 he took them to be those of Zephyrus betidae, until they made a round, 

 red-brown, very dense cocoon on the ground. The pupa, he notes, had a 

 very thin shell, and the position of the mouth-parts stood out strongly. 

 He concluded that it was likely that some species of fly or wasp would 

 be developed from them, but in the spring, Phalaena limacodes appeared, 

 the male smaller and darker than the female. He adds that much 

 as the larva superficially resembles that of Z. betulae, it differs in 

 having a short, lateral, bright, rose-red stripe on each segment, 

 whilst the yellow, oblique, lateral lines are wanting, and the ventral 

 area is covered with many pale yellow spots. 



p. 470. Anthroceramtdicayinis,Bdv. Under the name of A. ditbia 

 ab. confluens, Oberthur figures a form of this insect. He writes : 

 " Anthrocera dubia ab. confluens, Oberth., "Etudes," etc., xxth. livr., 

 pi. viii., fig. 147." This figure shows a specimen of medicaijinis, Bdv., 

 with an extra spot between 3 and 5, and a slight inner extension of 

 5 towards the extra spot. From Vernet-les-Bains. 



ERRATA. 



p. 15, line 23, for " Catocalia " read " Catocalid." p. 183, line 2 from bottom, for 

 " tengstormi " read " tengstromi." p. 416, line 13, for " paludis " read "palustris," 



