( 4G2 ) 



b;isal. rod-like, portion either iimueiliately nui>; ont iiitii the saw ff. -i'-V), or has to 

 cnrve a little dorsally to join the saw, the latter beinfr a little more dorsal than 

 the former; this variation is independent of locality. 



a. P. machfionjinvus from Great Britain. 



The denticulate portion of the harpe is generally longer than in machaon 

 machfjon from Germany and in machaon xjili^/riis from 8onth ICnrope and Asia Minor. 

 F. 39, 40, and 41 are taken from three British sjieoimens, iind represent the amount 

 of variation found by ns in onr series. The proportion of the length of the saw to 

 the entire harpe is in the three speeimens 19 : 30, 15 : 30, and 13 : 30 ; the variation 

 in the length of the denticulate jiortion amounts therefore to almost oil per cent, 

 of the length of the saw figured in f. 141. 



The U71CUS of the British specimens (f. 44, dorsal view) is nearly always slenderer 

 than that of the Continental individuals (f. 45), and agrees very well with that of 

 the Japanese subspecies (spring and summer broods). 



b. /'. macliaon machaon. 



In f. 45 the wncus of an individual from Switzerland is represented to show the 

 divergence from f. 44. 



c. P. machaon gphi/nis. 



The harpe of this southern subspecies as well as the harpe of machaon machaon 

 agrees on tlie whole with f. 40, but in some specimens from Asia Minor and 

 Palestine the denticulate portion is remarkably short. F. 42 represents an extreme, 

 the proportion of saw to entire harpe being 11 : 30 : the saw is in this individual 

 from Syria more than 70 per cent, shorter than in the British specimen represented 

 by f. 39, an amount of variation which  is higher than we anticipated. The harpe 

 of a Palestine specimen, as drawn in f. 43. is abnormal in so far as the saw does 

 rise gradually, not abraptly. 



d. /-■. machaon hipjmo-ates from Japan. 



Besides the uncus mentioned before, we do not see any difference between the 

 genital armature of this remarkable form and the European machaon. Tlie spring 

 and summer broods, though so consjacuously different in size and pattern, also do 

 not exhibit, to our knowledge, any distinguishing character in the organs in question. 



As tlie swallow-tails found in the ^Vicken Fens, near Cambridge, are doubtless 

 one species, the variation of the species in the length of the denticulate jiortion of 

 the harpe, or, in other words, of the prehensile part of the harpe, amounts to nearly 

 50 per cent. As further the Syrian specimens do not differ in the wing-pattern so 

 much from British sj)ecimens as in (Jermany and in Syria the individuals of the first 

 brood very often differ from those of the second brood, and as a line of separation 

 between the characters of British and Syrian specimens is altogether absent, we 

 have also to admit that f. 39 and 42 are taken from the same species, i.e. that 

 the variation of the preiiensile organ amounts to 73 per cent. 



Onr conclusions arc : (1) The genital armature of machaon does not afford any 

 characters by which the various geographical races, distinguished especially by 

 differences in the wing-markings, can be constantly differentiated from one another. 

 The iiarpc of the British form has, liowever, on the whole, the longest jirehensile part. 



