272 



BRITISH LEPIDOPTERA. 



A. Hindwings with only one inner marginal nervure. Antennae 



of male pectinated to the tip 



1 . Nervure III! arises from the upper corner of the discoidal 



cell . . . . . . . . . . . . . . Agliiinle. 



Wings broader, the apices of forewings rounded . . Loepa. 

 Wings narrower, the apices of forewings acute 



Central spots transparent . . . . . . . . Salassa. 



Central spots opaque . . . . . . . . Aglia. 



2. Nervure III! arises above the radius . . . . . . Automerin^. 



Forewings falcate . . . . . . . . . . s . Gamelia. 



Forewings with outer margin entire 



Hindwings above marked with large ocellated spots . . Automeris. 



Ocellated spots imperfect . . . . . . . . JJollipe. 



Hindwings without ocellated spots. Upper row of 

 pectinations of antennae of the male more than half as 

 long as the lower row. Nervure III! and IV 2 of hind- 

 wings forked 



Nervure IV 2 of upper wings arises from the middle 

 of the closed discoidal cell 



Nervure IV 2 arises near the upper corner of the 

 closed discoidal cell . . . . ... 



Nervure III and IV T of hindwings separate 



Smaller species with falcate forewings 



Upper row of pectinations of antennae of male less 

 than half as long as lower row 



Antennae of male with a single row of teeth 



B. Hindwings have two inner marginal nervures. Antennae of 



male pectinated for two-thirds of length only 



3. Nervure IIIi arises from the upper corner of discoidal 



cell or immediately beyond 



Nervure III! arises beyond the end of cell 



Outer margin of forewings longer than inner margin 

 Outer margin of forewings shorter than inner margin 



Medium-sized species ; The sexes similar 



The sexes dissimilar 

 Nervure III! arises from upper corner of discoidal cell 

 Large species : The abdomen scarcely longer 



than hindwings 



The abdomen exceeds the hindwings 



Dyar criticises (Ent. Rec, x., pp. 36 — 37) Grote's arrangement, 



and offers the following (based on larval characters) as a more natural 



grouping (compare anted, pp. 270 — 271) : 



I. A single dorsal tubercle on 9th abdominal segment, 

 stage present : 



Ormiscodes. 



Than ma. 



Plate ia. 

 Hylesia. 



Endyaria. 

 Dirphia. 



ClTHERONIIN/E. 



Anisota. 



/ Adelocephala. 

 I Sphingicampu. 

 Sys sphinx. 



Eacles. 

 atheroma. 



the primitive first 

 Citheronia. 



1. Anal plate tubercular ; tubercles unarmed 



2. Anal plate smooth ; tubercles covered with modified 

 urticating setae — 



a. Tubercle i with long shaft . . . . . . Automeris. 



b. Tubercle i with short shaft . . . . . . Hemileuca. 



II. No single dorsal tubercle on 9th abdominal segment, 



anal plate tubercular : 



1. Primitive first stage present*, spines unequal; 



mature tubercles atrophied . . . . . . Aglia. 



2. No primitive first stage, spines subequal — 



a. A dorsal tubercle on 8th segment . . . . Attaats. 



b. No unpaired dorsal tubercle . . . . . . Saturnia, 



Dyar objects floe, cit.) to Grote's division of the Attacids into 

 two groups on a single neurational character — the degree of approach 

 that IV 2 has made to the apex of cell, and points out that the 

 variation in the position of the nervure does not lend itself to the 

 establishment of a dichotomous division. On the other hand, if 

 the number of anal nervures in the hindwing be taken as a basis — 



* We are not (mite sure what Dyar means by "primitive first stage" in 

 Aglia. Poulton's figures I and 3 (Trans. Ent. Soc. London, 1888, pi. xvii) 

 appear to us to show an exceedingly specialised first stage. 



