92 ENTOMOLOGICAL NEWS. [March 



mounted. The errors resulting are obvious. To illustrate, 

 try the following experiment : Hold a sheet of ordinary note 

 paper before the eyes, flat surface parallel to the face, but 

 slightly bent, as the last ventral segment usually is over the 

 convexity of the venter. Its upper edge, if evenly held, will 

 be truncate. Tip this edge very slightly toward the face and 

 it becomes distinctly incurved ; tip it away and it becomes 

 ''broadly rounded." ^N^ow fold the sheet along the middle 

 line, spread it nearly but not quite to its original position, and 

 repeat the experiment. This represents some very ordinary 

 conditions in the last ventral segments. Tipped very slightly 

 towards the face it becomes deeply angularly emargiuate ; 

 tipped away it becomes strongly angularly produced at middle, 

 with a concavity on either side. These w^idely different ap- 

 pearances can all be produced by variations in the position of 

 a rectangular sheet of paper. Now if the sheet be cut into 

 some of the various shapes which the last ventral segment really 

 does assume, and its position varied as above, still more re- 

 markable changes are readily produced. To such causes in 

 part are due several of Mr. Gillette's species, and perhaps the 

 effects observed by Osborn and Ball, and mentioned on p. 737 

 of Gillette's paper. 



Under these circumstances individual variation is greatly 

 accentuated. The necessity of correct observation, then, is of 

 first importance. Next in importance measurements should 

 be made directly from the specimen. Measurements made 

 from camera lucida drawings, especially if the drawings are 

 taken on a flat surface, are very inaccurate, sometimes showing 

 almost a specific difference between the two sides of the draw- 

 ing. Delicate calipers or an eye-piece micrometer should be 

 used directly, i 



The forms remaining in group I, as above mentioned, I 

 would arrange as follows : 



1. SMARA.GDULA, representing the only specific type in the group. 



a. variety clypeata. 



{livingstonii). , 



b. variety aureoviridis. 



(mcisa). 



c. variety obtusa. 



[pergandei, denticuU and unico'or). 



