362 The Pedigree of Insects 



tion of typical parts of the primitive jaws rather than 

 by such startling transformations as have produced 

 the sucking-tube of Moths. On the other hand the 

 union of the first hind-body segment with the thoracic 

 capsule is a pecuHar feature unknown in any other 

 insects ; and in the higher families this specialisation 

 of the fore-body is accompanied by the remarkable 

 constriction of the succeeding abdominal segments 

 which gives rise to the characteristic "waist" of 

 wasps and ants. In the larval stage a transition can 

 be traced from the sawfly caterpillars, closely like 

 those of moths, to the legless maggots of ichneumon 

 flies and bees, but, throughout the order, the larval 

 head is well-developed. Among the Diptera the 

 larvae have without exception lost their primitive 

 limbs, though some possess special sucker-feet or 

 " pro-legs," while in the extremely degraded maggots 

 of the Muscidse and allied families the head region 

 is represented only by a protrusible proboscis with 

 hooks. The body of the imago is always highly 

 specialised. The mandibles are present as piercers 

 or altogether suppressed, while the sucker is so far 

 modified that it is a disputed point whether it is 

 composed of the first or the second maxillae ; the 

 hind-wings are reduced to small, stalked knobs. 

 From the lower families of Diptera — the Crane-flies 

 for example— to the higher — such as the Muscids — 

 a condensation of the body-segments can be traced 

 accompanied by a simplification of wing-neuration, 

 and a modification of the feelers from a typical, many- 

 jointed, thread-like form to the short antenna of a 

 blow-fly with its three specialised segments. 



The Hymenoptera and Diptera are not nearly 

 related to each other and either order is so isolated 

 that it is hard to suggest what their affinities to other 

 insects may be. There is not much evidence from 



