418 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.54. 



order., it should not on that account he arranged in such order^ hut, 

 on the contrary^ form a distinct one. 



The metamorphosis of the Strepsiptera is classed as hypermeta- 

 morphic, begimiing with larviparous reproduction of free living 

 hexapod larvae, which are conveyed by various means to the larvae 

 of their future hosts. These hexapods after beginning the parasitic 

 existence distend and become legless, and each succeeding molt makes 

 the female more degenerate, while the males undergo a transforma- 

 tion of specialization. Both sexes exsert the head and thorax from 

 the abdomen of the host as larvae, and the male pupa is formed 

 within this last larval skin. The female remains imprisoned within 

 the last larval skin and has no pupal stage, remaining absolutely 

 larviform. 



The larviparous reproduction occurs in the family Micromalthidae 

 of the Coleoptera, in Hemiptera, in Diptera, and elsewhere in in- 

 sects. There is nothing on this score to associate the Strepsiptera 

 to any one of these orders. The hexapod larva of the Strepsiptera 

 has its counterparts in the triungulin of the Meloidae (figs. 2, 3), 

 the triungulinid of the Rhipiphoridae (fig. 4), the planidium type 

 of larvae in the Hymenoptera, and especially the first larvae of the 

 Dipterous family Cyrtidae. The larvae of Pterodontia flavipes Gray 

 of the Cyrtidae are parasitic in spiders. They look more nearly like 

 a Strepsipterous triungulinid than any of the others but are distin- 

 guished by the absence of legs. The internal chitinous structures of 

 the Strepsipterous larvae and the backward pointing mandibles are 

 points of resemblance to the Diptera and of separation from the 

 Coleoptera. 



However, no other insects have a metamorphosis which is similar 

 throughout to the Strepsipterous type, and we have but one type in 

 the entire group. Metamorphosis can not link the Strepsiptera to 

 either the Diptera or the Coleoptera because the structure is not 

 similar to either of these orders. 



Rule IV. Where the genera which compose an order have invari- 

 ably one kind of metamorphosis., no insects that vary from it in that 

 circumstance should he placed in it, unless they exhihit a perfect agree- 

 ment with it in characters. 



The genera of Strepsiptera have invariably one type of metamor- 

 phosis. They can not therefore be placed with any other order 

 which has a different type of metamorphosis. This precludes their 

 being placed in the Coleoptera, Neuroptera, Diptera, or Hymenoptera, 

 with all of which various authors have associated them. Certain 

 Coleoptera have a type of hypermetamorphosis with points of simi- 

 larity, but these Coleoptera by virture of their characters, under 

 Rule II, remain Coleoptera. The Strepsiptera could only be ar- 



