108 MORPHOLOGY, ETC., OF THE MICROPTERYGID.E, i., 



the alar trunk; and, thirdly, that, in a number of families {e.g., 

 in the Satumiidce and some Butterflies*) the radius is likewise 

 split back to the alar trunk, and may even arise as five separate 

 trachea? (Antherea eucalypti); yet noWody questions that these 

 all belong to the radius. 



If Professor Comstock's interpretation be correct, we are 

 bound to ask, how is it that it is in the ancient Micropterygidce 

 that this specialisation (i.e., the fusion of Cu and 1 A) reaches its 

 highest expression; and how is it that, in the most highly special- 

 ised of all Lepidoptera, viz., the Satumiidce and the Butterflies, 

 we find a retrogression to what, on this view, must be the nearest 

 approach to the primitive type, viz., that in which Cu and 1A 

 arose separately from the alar trunk 1 This question is unan- 

 swerable, except by the admission that it is the higher families 

 of the Lepidoptera which show the greatest splitting back of the 

 traehese, while the original condition is preserved more completely 

 in the older families, and especially in the Micropterygidce. In 

 other words, the cubitus is three-branched, and the true lAis 

 the first of the quite separate anal group of trachea 1 , lying in its 

 natural position, posterior to the anal furrow. j 



It should also be noted that there are already three separate 

 anal traehese recognisable in the anal group, without the sup- 

 posed 1A of Comstock's interpretation. If Comstock is right, 

 then it is necessary to explain how the Micropterygidce (and most 

 Lepidoptera) come to possess four anal veins, whereas the older 

 Orders Mecoptera, Megaloptera, Planipennia, and Trichoptera 

 are admitted to possess only three. 



In Plate iii., fig.3 and Text-fig.2, the actual condition of the 

 primary cubital fork in the pupa of Eriocrania is shown. It 

 will be seen that the line of the main stem of Cu is continued 

 beyond the fork, not by Cu,, as we might have expected, but by 

 Cu. r From the fork itself, Ciij arches up at an angle to the main 

 stem, and then turns to run parallel to and above Cu 2 until it 

 again forks into Cu and Cu ]h . A pale band passing from M to 



* Also sometimes in the ancient Hepialidn ! 

 t The full proof of the limits of Cu in Lepidoptera will be given in Part 

 iii. of the "Panorpoid Complex." 



