SIR JOHN LUBBOCK ON THE ANATOMY OF ANTS. 143 



metathorax, as usual, and evidently correspond with the two pairs of thoracic spiracles of 

 other insects. The third pair is situated at the side of the so-called metathorax ; but in 

 no case whatever do we find among insects two pairs of spiracles on one segment. Such 

 an arrangement would be contrary to the whole plan of organization of the Arthropoda. 

 It is obvious, therefore, that the third pair of spiracles corresponds to that which in other 

 insects lies between the thorax and the first abdominal segment. Burmeister, as we 

 have seen, remarks that certain Hymenoptera " distinguish themselves from all other 

 orders " in having a pair of spiracles " on the metathorax itself;" but he supposes that 

 these correspond to the spiracles which are ordinarily situated between the meso- and 

 metathorax, overlooking the fact that these spiracles also exist as usual. It seems clear, 

 therefore, that the portion of the body posterior to the third pair of spiracles really cor- 

 responds to the first abdominal segment in ordinary insects. 



Nor are the respiratory organs alone in pointing to this conclusion. The internal 

 chitinous appendages clearly divide the thorax into four portions ; and I think it may 

 be said that the thorax contains four ganglia, though the last (PI. XL fig. 2, G 4 , PL XII. 

 fig. 2) is certainly not large. 



The Prothorax. 



The upper part of the prothorax, or pronotum, is formed in Lasius jlacus by a single 

 arched chitinous plate (PL XL figs. 1, 1, & 6, B), which slopes downwards from its posterior 

 border towards the head, where it forms a sort of keel (PL XL figs. 1 & 2). Seen exter- 

 nally and from the side, its lower border appears to join the upper edges of the propectus ; 

 but a transverse section (PL XII. fig. 4) shows that this is not so, but that the propectus 

 is continued for some distance beyond the lower margin of the pronotum, and is then 

 connected with it by a membrane which passes from the upper margin of the propectus to 

 the lower one of the pronotum: The propectus tapers in front (PL XL fig. 5), terminating 

 on each side in two teeth, which lock into two corresponding teeth (PL XL figs. 6 & 7, 

 and PL XII. fig. 1, X) or processes at the back of the head. The propectus is divided 

 into two plates (Plate XL fig. 2, C & T), one anterior and one posterior, which, moreover, 

 are divided into lateral regions by a central ridge. The anterior plate of the propectus 

 has in front a deep bay or depression, at the two horns of which are the above-mentioned 

 teeth or processes. Each region of the anterior division of the propectus has therefore 

 roughly the form of a triangle with arched sides. The posterior division of the pro- 

 pectus is elliptic in form, and not so large as the anterior division, to which it is firmly 

 attached. 



The propectus is therefore attached to all the surrounding chitinous plates by flexible, 

 though tough, membranes. It hangs, indeed, something like the under body of a carriage; 

 xnd from the fact that the anterior horns of the prothorax interlock with the posterior 

 arocesses of the head, if the propectus is turned round it carries the head with it. On 

 ;he other hand, if the head be retracted, the posterior processes of the head, from their 

 position with reference to the anterior horns of the prothorax, prevent the head of the 

 nsect from being turned round against its will. 



