of the Pyralidiiia of the European fauna. 431 



never be determined except by actual consideration of 

 the particular circumstances. Nor can it be said before- 

 hand what characters are likely to be good ; perhaps 

 the most suspicious are tufts of hairs, especially when 

 developed as secondary sexual characters, when they are 

 often unreliable. 



In the use of generic names I have followed the now 

 generally received practice of adopting the generic name 

 under which a species of the genus was earliest de- 

 scribed, except where such name has been preoccupied 

 in a different sense by another author ; subsequent 

 limitations being accepted so far as they restrict the 

 meaning of a generic name in accordance with my 

 definition of the genus. The misuse of some older 

 names is largely due to an indiscriminate following of 

 Treitschke. To give one or two conspicuous instances, 

 the genus Botijs was founded by Latreille to include two 

 species only, now passing as Lythria imrpuraria and 

 Hydrocampa nymphcBata ; it must be long since either of 

 these species was included by any writer in Botys, but 

 clearly one or other must be the actual type ; I hold it 

 to be purpuraria. Scopida, Schrk., was founded to 

 include siratiotata and dentalis, and is a synonym of 

 Nymphida. Both these names were subsequently used 

 by Treitschke in a quite different sense, for which there 

 is no authority. Ahicita and Fterophorus are also 

 instances of generic titles much abused. In some 

 instances a generic name has been orthographically 

 wrongly written in the first instance ; I have concurred 

 in the prevalent view that, in the interests of perman- 

 ence, such an error is not to be corrected, as it opens 

 up an unending possibility of confusion, except where it 

 is a mere printer's error for which there is evidence (see 

 Psammotis). In specific names the necessity for absolute 

 literal permanence does not exist to the same acute 

 degree, and corrections may, I think, be sometimes made 

 here, when the error is slight and the intention of the 

 writer obvious. But I hold that it conclusively follows 

 from this that, if a generic name is not liable to modi- 

 fication in the slightest degree, then any original diffe- 

 rence, even one of the termination only, is sufficient to 

 constitute two names distinct for separate use. Indeed, 

 as it has hardly ever been proposed to alter the termina- 

 tion of any generic name, there is no probability of con- 



TRANS. ENT. SOC. LOND. 1890. — PART III. (SEl'T.) 2 G 



