( 539 ) 



Abdomen striped with lilack and yellow, antenna pale 

 tawny ; or abdomen black, dotted with yellow, only 

 one row of dots on each side of the tergites, antenna 

 black Troilus Group. 



Abdomen entirely yellow beneath and at sides ; or all 

 black, except a row of yellow dots situated laterally 

 on the sternites, there being no dots on the tergites ; 

 or the abdomen black (? ¥), with a yellow line 

 at the lower edges of the tergites, the hindwing 

 of these females bearing a regular row of yellow 

 snbmargiual spots at least on underside . . Thoas Group. 



V. Machaon Group. 



Antenna reaching to two-thirds of cell of forewing ; club thick, obtuse, the 

 last segment very short, the preceding three or four more than twice as broad 

 as long. Abdomen either with yellow longitudinal bauds, or black, with rows of 

 yellow spots, two rows on each side of the tergites. Tail nou-spatnlate. Basal 

 half of hindwing yellow or black, without the large V formed by black bands in 

 the Glaucus Group. Cell of hindwing not widened. Harpe of male saw-like, tiie 

 distal portion being a denticulate ridge and the pro.ximal portion being less elevate, 

 non-dentate, subeylindrical. In female on each side of vaginal orifice a long 

 three-cornered flap which is dentate at the edges ; proximally of these flaps 

 a continuous, slightly elevate, ridge from side to side ; behind the orifice a 

 membranaceous tubercle clothed with extremely small hairs. 



The anal eye-spot, composed of the marginal and submarginal spots, is better 

 developed in this group than in the Glaucus Group. The origin of the ocellus 

 from the two sjjots mentioned is well illustrated by P. daunus and allies ; also 

 iu the present group the component submarginal and marginal sjjots remain 

 occasionally sejiarate. 



With the exception of the Asiatic P. xuthu.s, which stands apart, the species 

 of the Machaon Group are all very closely allied to one another. P. indra is the 

 only one which is recognisable by the genital armature, all the others being identical 

 in structure, as far as we know. The larvae also not presenting any very trenchant 

 characters, one might well ask if P. machaon, the dimorphic P. bairdi, the variable 

 P. polyxenes, the comparatively constant P. ::cUcaoii and P. nitra are really 

 specifically distinct from one another. Judging from the structure and pattern 

 aloue, one might be inclined to regard all these insects as individual and 

 geographical forms of one species. However, one important fact is known which 

 speaks entirely against this assumption. Papilio bairdi and P. pobjxen.cs asterius 

 occur in the same canons in Colorado, but keep perfectly separate, the one living 

 as caterpillar on Umbclliferae, the other on a Composite plant {Artemisia dracunca- 

 loidu.i). If these two insects exist iudej)endently side by side — i.e. are true species — 

 there is no reason to treat P. zclicaon otherwise than as a species by itself. And 

 P, machaon aliaska, which occurs as far south as Oregon according to the two 

 Edwardses, must also be considered independent of P. bairdi f. oregonia in spite 

 of the close general resemblance in pattern. We doubt if P. nitra is more than 

 a northern form of P. bairdi. 



35 



