nr R. .1. TILLYARD. 195 



most probable cause of degeneration \vould be the using-iip of 

 the veinlet-chitin in the formation of the accumulated masses or 

 tubercles in which the hairs are inserted. Thus Archipanorjta 

 would illustrate an intermediate stage in the evolution oi a fvimj 

 with a hairy inemhrane from an ancient net-veined or FaUtodicty- 

 opterous wing- type, and would give us the clues both as to the 

 method whereby an open-veined wing, such as tliat of the J^epi- 

 doptera or Diptera, has been evolved from a closely- veined type, 

 and also as to the method whereby the hairs (or, later on, the 

 scales) become seated upon the membrane as well as upon the 

 main veins. 



(3) Hairs upon the wing-nieinhraue. — In those places where 

 the structure of the tubercles of the wing-membrane can be 

 most clearly made out, they are seen to be very closely similar 

 to those upon the wing-veins, both in size and form. We can 

 distinguish an outer raised rim (indicated by the outer circles 

 in Plate viii., fig. 6), and an inner depression or hollow (indicated 

 by the inner circles in the same figure). There can be no doubt 

 that these structures represent the tubercular bases of insertion 

 of stiff bristle-like hairs; for the insertions of such hairs in the 

 wings of recent Planipennia are exactly similar to them. I 

 propose to call these hairs macrotrichia. 



Besides these tubercles, a careful examination of the wing- 

 membrane shows tliat it is covered all over with minute pits, 

 mere depressions without any visible raised rim, and of a size 

 so small as to be rather easily confused with the grain of the 

 rock on which the fossil is impressed. These, I consider, repre- 

 sent the insertions of minute hairs, several times smaller, and 

 far more numerous, than the macrotrichia. These hairs may be 

 termed micr atrichia. In most parts of the wing, they run in 

 closely-set parallel lines, passing alike over ridge and hollow. 



The structure of the pterostigma differs from that of the rest 

 of the wing-mem i)rane only in lacking the polygonal ridges, and 

 in being of a single thickness throughout. This makes it appear 

 very smooth in comparison with the rest of the wing. On ex- 

 amination under a moderate power, both macrotrichia and micro- 

 trichia can be seen to be pi'csent upon it. 



