36 



BULLETIN 491, U. S. DEPARTMENT OE AGRICULTURE. 



but slightly rounded, bearing 18 or 20 lobes, as illustrated in figure 6, a. Pos- 

 terior stigmal plates very evident, not appreciably elevated or depressed, ap- 

 proximate, each with three well-chitinized slits pointing toward those of the 

 opposite plate as illustrated in figure 9. Slits proportionately longer and 

 narrower than in either of the other instars. The fine bristles about margin 

 of stigmal plate arranged in four sets, relatively finer and shorter than in 

 the second instar. Anal opening on venter of last segment, with a rounded 

 protuberance on either lateral side, the entire area including these protu- 

 berances elevated to form a distinct tubercle armed about its base by a 

 concentric area of minute spinules. Body of larva glabrous, except for areas 

 of microscopic spinules arranged as follows : Band on Newport's segment cov- 

 ering the anterior two-thirds of the dorsum but widening ventrally to cover 

 the venter ; narrower bands extending about the anterior margin both dorsally 

 and ventrally of the second, third, and fourth segments ; fusiform bands 

 composed of broken rows upon the venter only of segments 5 to 12. There 

 are no lateral spinulose areas on segments 5 to 12. 



Fig. 8. — The melon fly : Third-instar larva, a, Lateral view of entire body ; li, dorsal 

 view of anterior end ; c, d, lateral and ventral views of same. Enlarged. 

 (Original.) 



mTEATION OF LARVAL INSTARS, 



The larva of the melon fly passes through three instars before be- 

 coming full grown. The number of hours passed in each instar has 

 been obtained for the first 12 larvae recorded in Table VI. These 

 larvae were transferred daily to fresh pieces of ripe papaya pulp, 

 and were therefore given the best combination of food and tempera- 

 ture for rapid development known to the writers. The majority of 

 these 12 larvae passed about 1 day in each of the first two instars and 

 about 2^ days in the third instar. An examination of the number 

 of hours for each instar recorded in Table VI shows that variation 

 may occur at times when no apparent cause can be assigned and that 

 a larger number of hours passed in one instar may be followed by 

 fewer hours than usual in the succeeding instar. As shown by the 

 data in Table VI, larval development may be completed in as few 



