THORAX. 



265 



" It must not be imagined, however," obsei*vesMr. MacLeay, 

 " that the pieces of the thorax mentioned in the above table 

 are all perfect and distinct in every insect. Pieces of the 

 thorax may disappear, being evanescent, owing to the great 

 developement of the contiguous segments, or by being con- 

 fluent or soldered together with the next adjoining pieces. 

 To know the pieces which are thus lost, it might be thought 

 that, on comparing the larva with the j)erfect insect, the posi- 

 tion of the stigmata (spiracles) ought to afford some clue ; 

 but in truth these are unsafe guides, as it is well kno\\Ti that 

 the situation of the stigmata in the perfect insect varies very 

 generally and considerably from what it was in the larva." 



Tigs. 118, Pronotum—119, Mesonotum— 120, Metanotum of a Dyticus. In 118 the subscg- 

 ments are confluent. In 119 and 120 the alternate subsegments are (lotted to show their 

 extent. 



We have indeed found a clue for getting out of the difficulty 

 above mentioned ; but from the little attention hitherto paid 

 to the subject, another difficulty of a not less formidable, but 

 of a far more philosophical kind, has presented itself, namely, 

 that of proving, by careful study, the relations of the several 

 parts in the different orders, and the variations they are 

 subject to. 



The total number of parts in the thorax, according to 



A A 



