4 A. WATSON 



species of which is an African Thyretid. Travassos (1943), Watson (1971; 1973) and 

 Kiriakoff (1973) have pointed out the incorrect family placement of Automolis by 

 earlier authors. The need to clarify the taxonomy of this heterogeneous assemblage 

 of Arctiid and Ctenuchid species has been evident for some time (Seitz, 1921 : 365) 

 and has been restated by Forbes (1939 : 192), Travassos (1943) and Watson 

 (1971; 1973). 



Blest (1964) has shown that some of the species discussed in the present work 

 exhibit aposematic coloration and are unpalatable to certain predators. The lack 

 of expected Miillerian concordance in colour-pattern discussed by Blest (1964) is 

 shown here to be an illusory anomaly (see remarks on mimicry) . 



The ultrasonic sounds produced by many nocturnal Arctiidae and Ctenuchidae 

 have been demonstrated to act as aposematic signals (Dunning, 1968). The external 

 structure and sound production capabilities of the tymbal organs in the genera 

 dealt with in this paper are discussed. 



Colour terms used in this paper are taken from Kornerup & Wanscher (1967). 



SUPRA-GENERIC CLASSIFICATION 



Since Forbes (1923), most authors have grouped together the Arctiidae, Agari- 

 stidae, Ctenuchidae, Hypsidae, Noctuidae and Nolidae in one superfamily, the 

 Noctuoidea. Brock (1971) supported Forbes' (1923) additional inclusion in the 

 Noctuoidea of the Notodontidae. Below family level there has been much difference 

 of opinion (Kiriakoff, 1952). Forbes (1939) recognized two subfamilies of Arctiidae: 

 Arctiinae and Lithosiinae, and later (i960 : 15) added the Pericopinae and, tenta- 

 tively, the Hypsinae. Forbes' (i960) postulation that the Ctenuchidae [his 

 Euchromiidae] are probably not separable at the family level from the Arctiidae, 

 is supported by the fact that tymbal organs (p. 7) commonly occur in both nominal 

 families although they may be vestigial in day-flying species of Ctenuchidae. The 

 adoption of such proposals, however, should await a comparative study of the 

 included genera. A subfamily classification which reflects the degree of similarity 

 of only the type-genera and a few of its allies is not a great improvement on that 

 existing now. 



All the genera surveyed in the present work can be placed in the Arctiinae sensu 

 Forbes (1939), the equivalent in Seitz (1918 : 231) of the combined Phaegopterinae, 

 Micrarctiinae, Spilosominae, and Arctiinae. At the tribe level, these genera can 

 be included in the Phaegopterini, as defined by Forbes (1939 : 192), and below 

 tribe level in the Eupseudosoma-gvoup of Forbes and Franclemont (1957 : 149) • 



SPECIES CLASSIFICATION 



The type-species of Automolis Hubner was designated by Kirby (1892 : 220) 

 as Sphinx meteus Stoll ([1781] : 109, pi. 347, fig. B) (type-locality: South Africa, 

 Cape of Good Hope). Zerny (1912 : 44) transferred meteus to Metarctia Walker, a 

 genus of Thyretidae, apparently as unaware of Kirby's type-species designation 



