BY R. J. TILLYARD. 679 



another); so that, if one attempts to draw the outUne of a cell 

 under the camera lucida, its exact shape and position depend 

 partly upon the direction of light. In order to show up these 

 areolets, it is necessary to use very oblique light. Consequently, 

 if the light be directed from the right side of the areolet, the 

 dividing lines between light and shade (which are the lines that 

 give us the rough outline of the cell) will lie too far to the right; 

 while, with light directed from the left, the}^ will lie too far to 

 the left. That this is so, I easily proved by drawing a typical 

 areolet or two, with two or more different lightings, when the 

 outlines of the cells did not wholly coincide. Therefore, in all 

 important cases, it was essential to make at least two drawings, 

 and to gauge the true outline of the cell as a middle position 

 between them. By this somewhat laborious method, I hope that 

 I have succeeded in obtaining a more accurate result of the 

 details of this wing than would otherwise have been possible. 



The general appearance of the fossil is seen from the two 

 photographs in Plate xlii. In Fig. 9,^ the light was directed on 

 to the fossil from the left, or basal side, at an angle of about 20° 

 to the plane of the wing. In fig. 10, the light was similarly 

 directed from the right or distal side. A comparison of the two 

 figures shows the strong convexity and concavity of the veins 

 very clearly. The following are the convex veins: — C (very 

 slightly), R + M to arculus, R beyond arculus, Ms, M4, A and 

 all its descending branches, and the distal side of the triangle. 



The following are the concave veins: — Sc, its apparent con- 

 tinuation beyond the nodus, the nodus itself, the subnodus and 

 its continuation Rs; Mj, M.,, M3, Rspl, Mspl, Cu, and the an- 

 terior and proximal sides of the triangle. 



It will be seen from this, that the plane of the triangle is not 

 horizontal, but inclined upwards distally, so that this area 



* The numbers of the figures and text-figures in this series of papers are 

 made concurrent from one paper to another. In No. 1, the figures in 

 PI. ix. were inadvertently left unnumbered; the forewing of Archipanorpa 

 maynifica should have been fig. 7, the hindwing fig. 8. Fig. 9 of PI. xlii. in 

 this paper follows on consecutivelj-. 



