— 276 — 



EXPLANATION OF THE PLATES. 



Plate VIII. 



a-c. Very young galls; the fissure is not yet closed (nat. size). 



d. The fissure is closing (nat. size). 



e. Fullgrown but still white galls; the hole is still open (nat. size). 

 ƒ. The holes are closed by a membrane (nat. size). 



Plate IX. 



Part of a Pandanus plant to show galls in situ, on the left empty galls, 



to show the discolored streak. (± I/2 of nat. size). 

 Plate X. 



Some leaves quite covered by old galls; in one the membrane is still 



closed (nat. size). 

 Plate XI. 



1. Cross cut through very small gall. The black spot, also with the other 



figures, is the young larva. The excavation is still entirely open 

 (real size 3 m.m.). 



2. Older gall, the excavation is closing (real size 12 m.m.). 



3. The same gall cut longitudinally (real size 17.5 m.m.). 



4. Older gall, the fissure is closed (cross cut, real size 15 m.m.). 



5. Old gall cut longitudinally to show the larval chamber (real size 40 

 m.m.). B: the breathing hole. 



6. Schematic view to show development of the gall. 

 Plate XII. 



- 1. Larva (real size 5 m.m.). 



2. Pupa, side view. 1 



OD * I • ( real size 4H m.m. 



3. Pupa, ventral view, j 



4. Anal segment of larva (real width l'/^ m.m.). 



5. Female (real length 4.5 — 5 m.m.). 



6. Chalcidid parasite, male (real length 3.5 m.m.). 



7. „ „ , abdomen of female (real length: abdomen 1%, 

 terebra 3 m.m.). 



(Pen drawings made by Kades, photographs and plate XI by author). 



