36 



BARNES AND McDUNNOUGH: CATOCALA 



readily show the points of distinction from similis Edwards and prcoclara Grote and Robinson, with which species several 

 authors have confused it. 



The species is not rare along the Atlantic Coast; it has been reported from as far north as the vicinity of Ottawa 

 Ontario, (Gibson, Rep. Ent. Soc. Ont. for 1912, p. 121) and extends southward, according to Beutenmuller, into Florida- 

 westward it ranges into Pennsylvania and Ohio and has been included by French in his synopsis of the Catocalse of 

 Illinois. 



Catocala andromedse (Guenee) 



Plate I, fig. 18; PL XXII, fig. 8 (clasper). 



Hypogram?na andromedce Guenee, 1852, Hist. Nat. Spec. Gen. Lep., VII, p. 36. 

 Catocala tristis Edwards, 1864, Proc. Ent. Soc. Phil., II, p. 511. 



This species has been generally known as tristis Edwards but Sir Geo. Hampsoil has recently called our attention to 

 the fact that GueneVs description of andromedos (a name heretofore unplaced) fits this species very well; we concur with 

 him in using GueneVs name, which has priority. 



The similarity of this species with the preceding in general type of maculation of primaries as well as in structural 

 characters leads us to believe that the two are correctly associated, although the early stages of both are unknown. 



The range of the species is practically the same as that of gracilis, although it is distinctly rarer. It appears to extend 

 farther down the Mississippi Valley than gracilis, as we have a single specimen in our collection labelled St. Louis, Missouri, 

 and several specimens from Harris County, Texas; concerning the authenticity of these labels, however, we know nothing. 



Group XVI 



Egg hemispherical, prominently ribbed. Larva with lateral filaments and a dorsal wart on the fifth abdominal segment, 

 which usually is very prominent, pointed backward, and conical. Male claspers symmetrical, apically strongly rounded. 



The larva? of all the species are not known but are probably all oak-feeders; in at least one species (ophclia) the 

 dorsal wart is considerably reduced in size. The spining of the hind tibiae is reduced to one or two spines between the 

 spurs, which in some specimens are entirely absent. 



Catocala herodias Strecker 

 Plate VIII, fig. 10. 



Catocala herodias Strecker, 1876, Lep. Rhop. Het., p. 121. Beutenmuller, 1903, Bull. Amer. Mus. Nat. Hist., XIX, p. 506; 1913, 

 Insec. Ins. Menst., I, p. 97. 



This rare species was for a long time only represented by the unique type specimen from Texas; of late years it has 

 been bred from larvse found on oak trees at Lakehurst, New Jersey, proving its validity to specific rank. Nothing, as 

 far as we know, has, however, been published concerning the larva nor have collectors made any attempts to secure the 

 full life-history of the species. 



Apart from these two widely separated localities we know of no other authentic records for the species; judging, 

 however, from the nature of the insect fauna of Lakehurst, we venture the prediction that the species will be found to 

 occur generally in the pine barrens of the Southern States. 



Catocala coccinata Grote 



Plate III, figs. 16-18; PI. XIV, fig. 1 (larva); PI. XV, fig. 20 (larval head); PI. XV, fig. 33, and 

 PL XVII, fig. 11 (segments); PL XXII, fig. 9 (clasper). 



Catocala coccinata Grote, 1872, Trans. Amer. Ent. Soc, IV, p. 6. Barnes and McDunnough, 1918, Bull. Amer. Mus. Nat. Hist., 



XXXVIII, p. 170. 

 Catocala coccinata var. circe Strecker, 1876, Lep. Rhop. Het., p. 121. 

 Catocala sinuosa Grote, 1879, Can. Ent., XI, p. 15. 

 Catocala coccinata var. chiquita Bartsch, 1916, The Lepidopterist, I, p. 3. 



The typical form of this species with rather even gray primaries is figured on plate III, figure 16; the type material 



