and on the Number of Abdominal Segments, in Insects. 177 



complete metamorphosis. These nine segments are, however, 

 but seldom conspicuous in the perfect insects of this division, 

 and only their dorsal half-segments : the number of ventral half- 

 segments is always less than that of the dorsal ones, although 

 both half-segments are equally developed in the larva. Of the 

 dorsal half-segments, the last, or even the last two, often disap- 

 pear at the a])ex of the abdomen, being retracted within its 

 cavity during the pupa state ; of the ventral half- segments, not 

 only the last or the last two are retracted at the apex during the 

 transformation to the perfect state, but also the first, and often 

 the first and second, disappear externally at the base of the ab- 

 domen, being usually pushed inwards so as to form a kind of 

 phragma between the thorax and abdomen. The first conspi- 

 cuous ventral arcus in the perfect insect is, therefore, not the 

 one corresponding to the first, but the ventral arcus correspond- 

 ing to the second or third dorsal arcus*. In this way it is to be 

 explained how the number of visible ventral half-segments in 

 perfect insects is often reduced to five, six, or seven, while the 

 number of the dorsal ones amounts to seven, eight, or nine. 



In counting the latter, we have always to begin with the one 

 which bears the pair of large spiracles, so characteristic already, 

 for the first abdominal segment, in the larva, however intimate 

 the union of that half-segment with the metathorax may be. It 

 is, for instance, so intimate in the Staphylinidse, that even Erich- 

 son, neglecting the stigmata, considered it for some time as a 

 part of their metathoraxf. It is still more intimate in the Hy- 

 menoptera aculeata, where the first begment is severed from the 

 rest of the abdomen by a more or less deep incision, and is 

 immoveably applied to the metathorax — constituting that part 

 which is called by MacLeay, Newport, and Westwood the post- 

 scutellum of the metathorax J. 



* Erichson, Archiv, 1848, ii. p. 61. 



t Erichson, Archiv, 1845, ii. pp. 80,81; Stein, Yergleichende Anatomic 

 d. Ins. p. 11. 



+ That the so-called postsciitellum of the metathorax in Hvmenoptera 

 aculeata is in reality the first dorsal abdominal segment, as contended by 

 Audouin and Latreille, is not only proved by the size and position of its 

 stigmata, corresponding to those of the first abdominal segment of the 

 larva (while the metathora.\ has nowhere any stigmata), but also by the 

 changes which take place in the segments during the pupa state. It is the 

 sixth segment of the lar\ a (second abdominal) which forms the petiolus, by 

 which what seems to he the whole abdomen is attached to the thoracic jior- 

 tion, the fifth (first abdominal) applying itself intimately during these 

 changes to the metathorax (Cf. the figures of Ratzeburg, copied by "NVest- 

 wooti, Introd. ii. fig. 86. 4 & 5). The three parts of the body, so consjri- 

 cuous in tlie "Wasp, do not, therefore, as generally believed, exactly corre- 

 spond to the head, thorax, and abdomen ; but the first to the head, the 

 second to the thorax -\- first abdominal segment, the third to the abdomen 

 — the first segment. 



