120 Journal New York Entomological Society. |v<>i. xl 



determine a longitudinal line in comparison with which the angle of 

 inclination of the sclerites should be measured was not easy. After 

 canvassing the external topography of the thorax carefully, we settled 

 upon the pleural articulation of the middle coxa for the anterior point 

 (PI. VIII, Fig. 2, /^) and the infero-lateral articulation of the thorax 

 and abdomen for the posterior, the two determining the base line bd 

 of Fig. 2. By comparing Fig. i * it will be seen that the two points 

 are so close together as to occasion difficulty in bringing them into 

 exact coincidence with the base line of the scale. 



2. Anatomical. — The anatomical sources of error were several, 

 (rt;) The articulations used to determine the base line are something 

 more than points in breadth. (/^) They are sometimes obscured by 

 hairs, (r) The sutures with which the indicator must be made 

 parallel are sometimes sinuous, and their general direction has to be 

 estimated. 



At first the skewness of the three lateral sutures and of the dorsal 

 carina were measured, but as the differences discovered were rather 

 less than the rather wide limits of probable error, only the first lateral 

 (humeral) suture was measured to the end, and that and the tilt of the 

 wing bases in the opposite direction are reported upon below. 



The diagrammatized photograph (Fig. 2) shows these angles': 

 ahc is the angle made by the humeral suture cb with the perpendicular 

 ab to the base line db, assumed to be parallel with the axis of the 

 body. This angle measured upon the arc x represents the degree of 

 skewness or inclination of that suture. 



cdb is the angle made by a line cd drawn through the wing bases 

 with the base linq bd, and is measured upon the arc :;. The wing 

 bases are assumed to have rested primitively upon the line ec parallel 

 to db. The specimen shown in Fig. i is nearly in position for meas- 

 uring this angle. 



The actual measurements were all made by the junior author upon 

 miscellaneous papered specimens. Each specimen selected was meas- 

 ured first upon one side and then upon the other, and after intervening 

 measurement of other species, was measured again, and then the aver- 

 age of all the measurements was taken. But one specimen was used 

 for each species and the sex was disregarded. Some of the first meas- 



* The specimen is off the base line in Fig. I, having slipped out of place just be- 

 fore this photograph was taken. 



