1913.] NATURAL SCIENCES OF PHILADELPHIA. 229 



Quadrangle with its upper side distinctly shorter than the lower 

 side. 



Ms separating from M1+2 at 3.5-4 mm. from the arculus, 2.5-1 mm. 

 from level of nodus. 



Postnodal cross-veins'* 17-19, instead of 21. 



Stigma bounding 3^ or 4| cells above, preservation imperfect in 

 most of these specimens. 



Ml and M2 separating at as much as 10.75 mm. from arculus, 

 instead of 9.67 mm. 



Beyond the ciuadrangle, i.e., between M4 and Cui, are seventeen 

 single cells (Nos. 3 and 4), or sixteen (No. 5), then two rows increas- 

 ing; hence very different from Prof. Cockerell's statement, p. 62, 

 which does not seem to agree with his figure (13 of Plate V). 



There is but one row of cells between Mi and Mia throughout, thus 

 agreeing with the description of P/i.(?) parallehis (I.e., p. 62) in 

 distinction from Ph. mirandus. 



Two specimens, Nos, 1 and 3, are sufficiently preserved to enable 

 a count of the cells between M2 and Rs to be made. From the margin 

 proximad these are, in No. 1, 7 marginal cells, then 3 rows for 3-4 

 cells, then 2 rows for 6 cells, then 1 row (which is consequently met 

 at 10-11 cells in from the margin), this single row comprising about 

 15 cells to the origin of Rs, three of these 15 being basad of the origin 

 of M2. Corresponding figures for specimen No. 3 are 5 (?) marginal 

 cells, then 3 rows for 4 cells, then 2 rows for 3 cells, then 1 row (which 

 is consequently met at 8 cells in from the margin), this single row 

 comprising about 20 cells to the origin of Rs, 5| of these 20 being 

 basad of the origin of M2. In B. M. No. I 8423, this single row 

 consists apparently of 15 cells, 4 of which are basad of the origin of 

 M2. In B. M. No. I 15,049, this single row consists of 15 or 16 cells, 

 3 and the greater part of the fourth of which are basad of the origin 

 of M2; No. 5 appears to be very similar. None of these arrange- 

 ments is identical with that described for Pfi. mirandus, where, 

 however, the single row of cells is met at 11 cells in from the margin. 



Prof. Cockerell gives two formulae for Ph. mirandus and Ph.{?) 

 parallelus, respectively (I.e., p. 62), these formulae being obtained by 

 taking "a section of the wing down from the basal part of the stigma 

 to Cu2, and count[ingl the numbers of rows of cells successively met 

 between the longitudinal veins .... commencing with that 

 between Ri and Mi." The numbers of cells so obtained will depend 

 on the exact position of the line of section, but Prof. Cockerell does 

 not specify this. Choosing, as the two points fixing this line, the 



