BY R. J. TILLYARD. 



617 



came lost at the apex. A much stronger probability is that 

 either R 4 or K, 5 lias been entirely suppressed, by lack of chitin- 

 isation, as in the ease of Crt2> and occasionally also M 2 . in a 

 number of Lepidoptera. Pending a full proof of what has 

 happened, I have labelled the unforked part of Rs, R 4+5 . 



The genus Protoplasa (Text-fig. 67) is remarkable in still 

 possessing the four branches of Rs, though R 5 appears some- 

 what crowded in between R 4 and M 1( The evidence of this 

 genus, then, reinforces that of Psychoda as tc the original four- 

 branched condition of Rs in the Diptera; and this condition will 

 therefore be adopted for the Archetype. Protoplasa also has 

 an extra cross-vein connecting M 3 with M 4) which is strongly 

 suggestive of an original Meeopterous or Paratrichopterous con- 

 dition, in which the apical forks were supplied with such excess 

 cross-veins . 



Text-Fig.67. 

 Forewing of Protoplasa Jitchii (fam. Tipulidae) after Comstock, but 

 with lettering altered. Lettering as on p. 535. 



Some light is thrown upon the condition of the anal part of 

 the wing in Diptera by a study of the Tipulidae. In Gyno- 

 plistia (Text-fig. 49) there is a clearly marked remnant of Cu 2 

 most strongly ehitinised basally, and fused with 1A (a much 

 more strongly marked vein), for some distance beyond the cubital 

 fork (cuf) . Vein 2 A is well developed, and is connected with 

 1A near its base by the inter-anal cross-vein. 



Turning next to the Braehycera, we may select Tab anus and 

 Bhyphus (Text-fig. 68) as representing two of the most archaic 

 wing-types for this Sub-order, though many other genera of 

 the families mentioned above, on p. 615, would do equally well. 

 The points to be noted are: — the reduction of Rs to a three- 

 branched condition, by means of the elimination of the fork of 



