142 Sffi JOHN LUBBOCK ON THE ANATOMY OF ANTS. 



Newport, in his article " Insecta," in the ' Cyclopaedia of Anatomy and Physiology,' 

 p. 920, says : — " at first we were inclined to Audouin's opinion, more especially on 

 account of what we shall presently find in Lepidoptera, in which the fifth segment, in its 

 atrophied condition, is as much connected with the thorax as with the ahdomen. On 

 further examination, however, we are satisfied that that portion of the metathorax 

 which is posterior to the incisure belongs to the third segment of the thorax." 



Fenger, in his " Allgemeine Orismologie der Ameisen " (Arch, fur Naturg. 1862, p. 315), 

 treats the thorax as consisting of three segments, and does not even hint at any dif- 

 ference of opinion on the subject. Mayr also, in his excellent ' Die europaischen For- 

 miciden,' p. 4, and Schenck, in his " Bcschreibung nassauischer Ameisen- Arten " (Jahrb. 

 des Ver. fur Naturkunde im Herz. Nassau, 1852), adopt the same view. Lastly, Forel 

 (' Fourmis de la Suisse,' p. 5) says that the thorax " se divise en trois segments, comme 

 chez tous les insectes : prothorax, mesothorax et metathorax." 



It would perhaps hardly be appropriate to refer to more general or condensed works 

 in which the thorax is stated to consist of three segments, as, for instance, by Owen 

 (' Lectures on Invertebrate Animals,' p. 193), Claus (' Grundziige der Anatomie,' p. 557), 

 Bolleston (' Forms of Animal Life,' p. cix), &c. ; for these eminent authors, thougb expres- 

 sing no qualification, perhaps only meant to describe a general, and not necessarily an 

 invariable, rule. 



Huxley, in the ' Introduction to the Classification of Animals,' p. 58, observes, with his 

 usual care and accuracy, that " three, or perhaps, in some cases, more, somites unite, and 

 become specially modified to form the thorax." 



Notwithstanding the high authorities who have adopted the opposite opinion, and 

 although the first appearance of the tborax seems to support their view, for my own part 

 I cannot but think that Batzeburg's opinion was correct. Packard (' Guide to the Study 

 of Insects,' p. 66) has given figures of the metamorphoses of Bombus, from which it seems 

 clear that the fifth segment of the larva forms the posterior portion of the thorax of the 

 perfect insect. Lacaze-Duthiers (Ann. des Sc. Nat. 1853, p. 231), Palmen (Zur Morph. des 

 Tracheensystems), and Beinhard (Berl. ent. Zcits. 1865) also advocate the same view. 



The position of the spiracles affords also strong evidence in support of the same opinion. 

 It is generally stated in works on the anatomy of insects that there are on the thorax 

 two pairs of spiracles, the first between the pro- and mesothorax, the second between the 

 meso- and metathorax. 



According, indeed, to Burmeister (' Handbook of Entomology,' p. 164), this is also the 

 case with the Hymenoptera, which " all possess four (spiracles) in the thorax, two of which 

 are upon the limits of the prothorax, between it and the mesothorax, and the other two 

 lie between the meso- and metathorax. In the Hymenoptera, in which the thorax con- 

 sists of a hard, horny case, and the segments are closely united together, the posterior 

 pair of spiracles lie upon the metathorax itself, whereby they distinguish themselves from 

 all the other orders." In fact, however, as may be seen from the accompanying 

 figures (PI. XL figs. 2, 4, 5), the thorax of Ants possesses, not two, but three, pairs of 

 spiracles. 



Tbe two first pairs are situated between the pro- and mesothorax and the meso- and 



