52 CLASSIFICATION OF THE ALEYRODIDJE. 
Realizing this, Riley and Howard made a careful description of the 
form. The material upon which their descriptions were based is 
preserved in the bureau collection, and the following includes a study 
of this same material. Unfortunately, a slight marking on the fore- 
wings of the species was not considered of importance by Riley and 
Howard, as it was no doubt overlooked by Curtis. This oversight 
has led Cockerell into the error of redescribing the form as iridescent. 
Now. however, that Riley and Howard's specimens, Cockerell's types, 
and many other examples are at hand for study, there is little doubt 
that all these forms are the same species. 
DESCRIPTIVE REMARKS. 
Pupa case. — The pupa cases representing the Riley and Howard 
material and those composing Cockerell's types, as well as numerous 
others, are identical in structure as follows: The case (PL XII, 
fig. 2) is surrounded by a row of bristle-like hairs inserted on the 
dorsum just within the margin. There are usually 13 of these to a 
side. One pair of long, prominent caudal hairs is also present. 
Margin entire, and within it all around (PL XII, fig. 7) is a row of 
large, simple, boat-shaped pores and the submarginal area is covered 
with very many minute, simple ones. Along the median thoracic 
region there are four pairs of small tubercled seta? and on the cephalo- 
lateral portion a pair of large compound pores of the type illustrated 
in Plate XII, figure 6. The abdominal region has along each side 
four large compound pores similar to the thoracic pair, and on either 
side of the vasiform orifice are two reduced ones. The vasiform 
orifice is nearly cordate and just cephalad of it is a pair of seta 4 , one 
on each side. The operculum is subelliptical, broader than long, and 
armed caudad with two prominent bristles. The lingula is conical, 
exserted, and possesses two pairs of bristles. Both operculum and 
lingula are minutely setose. 
On the underside (PL XII, fig. 3) the mouthparts, legs, and 
antenna? are distinct. The trophic tubercle is large, the labium quite 
distinct, while the seta? are usually carried in the form of a loop. 
The antenna? are of two segments, 0.33 mm. long, and the proximal 
segment is short, subcylindrical, and unarmed. The distal one is 
much longer and annulated. At its distal end a short unguis is 
present and at the base of this a fringed sensorium. The legs are 
short, thick, of two segments, and armed with a solitary curved claw 
(PL XII, fig. 4). 
There is considerable variation in the size of the pupa cases, and <>n 
this ground it might be possible to separate in'descens from cocois. 
Since, however, we have considerable difference in the cases of the 
Latter species, depending upon locality and food plants, it seems 
hardly justifiable to select the larger specimens and cm 11 them a differ- 
