CATALPA INSECTS. 667 
with a sudden jerk, they would jump tj a distance of several inches.* The seeds 
themselves and the whole contents of the pod were in every case in a decaying con- 
dition. The larvae were of very different sizes, some apparently being nearly full- 
grown, while others were evidently very young. 
Some ten days after the pods had been placed in a breeding jar the adult flies 
began to appear— minute yellow midges with dusky wings. From that time on 
through the fall occasional examinations of the pods showed larvae of all sizes still 
at work, many of the pods becoming entirely brown and dry before the middle of 
September. It was often a puzzling thing, in examining these pods, to find the 
points where the larvae made their exit, for the pupa state is passed under ground. 
Usually one, two, or three small orifices would be found, through which all the in- 
habitants of the pod must have issued. The manner in which this hole is made is a 
mystery. Examined from the inside, it shows marks of gnawings around its edge, 
and frequently spots are found where attempts to pierce the pod have evidently been 
made, but unsuccessfully. Yet as cecidomyid larvae have no horny masticating 
jaws, how have they made these orifices ? In pods which had evidently been attacked 
earlier in the season, while younger and tenderer, the holes were much larger and 
more abundant. Occasionally the pod will have become so dry that it will have 
cracked, and in such cases of course no other hole would be necessary. 
Larva.— Length, 3.25 mm ; greatest breadth (at middle of body), 0.7 mm . Color vary- 
ing from pale whitish to orange. Breast bone bright honey-yellow, .2l mm long, and 
Qgmm w ide at the fork. Integument very smooth, transverse ridges barely percep- 
tible, with a high power near the juncture of the segments. Sides of the body show 
the dividing line of the segments only as a slight notch, the junctures between the 
head and first thoracic segment and the eighth and ninth abdominal segments being 
most marked. Body apparently with fourteen segments. Antennae apparently 4- 
jointed ; first joint short and broad ; second joint short, much narrower than joint 1 ; 
third joint three times as long as joint 2, but of same diameter; joint 4 a mere point 
at tip of 3, apparently the continuation of a tube which can be seen in joint 3. Stig- 
mata very small, at the summit of almost imperceptible tubercles, the prothoracic 
tubercles and those upon the eighth abdominal segment being larger, more dorsal, 
and situated, the prothoracic at the front and the eighth abdominal at the hind border 
of its segment. The anal segment is very convex anteriorly, and almost truncate 
posteriorly, four or more small posterior projections being present. 
Adult male.— Length of body, 1.3 mm ; length of wing, 1.8 mm ; length of antennae, 
2 5 mm . Antennae, 26-jointed (2 x 24); joints pediceled, alternately single and double ; 
single joints each with a whorl of long hairs; double joints with a whorl of delicate 
short hairs preceding the long one. Head slightly gibbous above, the eyes meeting 
upon the summit. Cross vein given off at one-half the length of the subcostal, not 
very oblique ; second longitudinal vein nearly straight for three-fourths of its length, 
when it curves downward and reaches the margin of the wing somewhat beyond the 
apex ; third longitudinal vein straight for one-half of the wing-length, when it forks, 
the branches forming a right angle first, which is, however, lost by the almost imme- 
diate downward bend of the upper branch. General color, light yellow ; antennae 
fuscous, except basal joints, which are yellowish; legs somewhat shaded with fus- 
cous, and furnished with quite long whitish hairs upon the femora ; thorax above, 
with a long longitudinal dusky stripe on each side, also faintly dusky toward head ; 
abdomen light yellow, with many short whitish hairs; balancers and claspers yel- 
low, the latter dusky at tip; wings dusky, with a bluish iridescent appearance. 
*This habit is mentioned by Osten Sacken (Monogs. Dipt., i, p. 183) in the following 
words: "The larvae of several species, for instance, Cec. loti, Cec. pisi, and Cecid.'rumi- 
cis, have the power of leaping. Dr. Loew remarks that all such larvae belong to the 
subgenus Diplosis. Cec. populi Duf. performed its leaps by straining the horny 
hooks at the tip of its abdomen against the under side of the thoracic segments." (Du- 
four's Ann. Sc. Nat., 2d ser., xvi, p. 257.) 
