1864.] 563 



haps, as often happens, one mistake has given birth to another, and in 

 stopping one leak another has been opened. Tn the larva of the Gall- 

 gnats there are, of course, no legs. In the pupa the legs extend to the 

 tip of the abdomen, or even beyond it, and both legs and antenna?, as 

 is well known to be the case with all other Nemocerous Diptera, are 

 perfectly free and detached from the body. From repeated experi- 

 ments, I know that, in the case of the Willow gall-gnats, the pupa re- 

 mains in this state for a week and over, without the legs or antennfie 

 becoming any longer, before it transforms into the imago. It is likely 

 enough, indeed, that the legs and antenna? of the future pupa may be- 

 come partially visible under the very thin, delicate, and semi-transparent 

 integument of the larva, shortly before that integument is moulted ; 

 but still they will not then he free, as in the true pupa, neither will the 

 insect be as yet in the pupa state, properly so called, for that very reason. 

 I believe that it was from not attending to the distinction between ob- 

 tected legs and antennae, and free legs and antennae, in two radically 

 distinct states of the Gall-gnat, viz : the very mature larva and the true 

 pupa states, that the above quoted assertions took their origin. I have 

 probably examined at different times considerably over a thousand spe- 

 cimens of Willow (iall-gnats, some in the larva and some in the pupa state, 

 and I always found them either in one state or the other. Whereas if, 

 as Harris and Fitch assert with especial reference to a Willow Gall- 

 gnat, the change from the larva to the pupa state was gradually and 

 sloicli/ effected, as a newly-hatched chicken gradually and slowly ex- 

 changes its hairs for feathers, I certainly must have met with at least 

 a few specimens in the transition state, i. e. with legs and antennae free 

 but only ] or i or f as long as in the normal pupa. Authors are per- 

 petually forgetting, that Annulate animals pass from one state to another 

 only by suddenly moulting their skeletons, while Vertebrate animals 

 retain the same skeleton throughout, and pass from one state to another 

 by the slow and gradual accretion of new matter. Osten Sacken inci- 

 dentally remarks that the facts referred to above are "not mentioned in 

 the European authors." (See on this subject Z><}9^. N. A. pp. 184 — b; 

 Harris Inj. Ins. pp. 56G — 7.) 



Perhaps few things have contributed so much towards propagating er- 

 roneous views on such subjects as these, as the almost universal use of the 

 term "skin" as applied to the external integument of Insects, especially 



