1"916] Thoracic and Cervical Sclerites of Insects 41 



The spiracle does not occur in all the thoracic segments, 

 being absent in some cases. Whether the spiracles which often 

 occur between the thoracic segments belong to the immediate 

 anterior or to the immediate posterior segment, is uncertain 

 and at present a disputed question. What Hagen considered 

 rudiments of wings on the prothorax of Strepvsiptera and 

 agrionids, have proved to be merely flaps and are no longer 

 considered homologous with wings. Several fossil Palaeodic- 

 tyoptera have a pair of lateral wing-like appendages on the 

 prothorax, like rudimentary organs, but it is very uncertain 

 whether these appendages do or do not represent traces of 

 wings. 



Patten, '90, claims that each thoracic segment is composed 

 of two annuli. Using the nervous system of Scolopendra as a 

 type, he found in all arthropods studied, two cross commissures 

 in each neuromere, thus indicating the double nature of these 

 structures. In Acilius the median furrow between the cross 

 commissures was similar to that found between the successive 

 neuromeres. In Scorpio, the neuromeres are distinctly double 

 and in such forms as Julus, the cardiac ostia, arteries, tracheae 

 and legs show the double nature of the somites. Segmental 

 fusion is also indicated by two pairs of tracheal invaginations in 

 each segment of Acilius, by the bifurcated appendages of many 

 Crustacea and by the bifid maxilla of insects, the latter group 

 frequently having monsters with double pairs of legs. 



Patten's double segment theory based upon the reasons 

 cited above, does not coincide with the facts in all cases, and 

 hence is open to objections. Furthermore, what occurs in 

 Scolopendra should not be given too much weight, when con- 

 sidering the condition found in insects. In insects the last 

 abdominal segment is considered as the fusion product of sev- 

 eral neuromeres and yet it contains only two cross commissures, 

 whereas according to Patten's idea, there should be as many 

 commissures present as neuromeres. In Vermes the neuromeres 

 of each segment contain but two cross commissures, although 

 the segments are not double. The median furrow between the 

 cross commissures in Acilius, if there is a furrow, which is 

 doubtful, can hardly be similar to that between the successive 

 neuromeres, for the latter are separated by a considerable dis- 

 tance. In such forms as Julus, Patten found the double nature 



