Constitution of the Body in the Blattidse. 231 



on the other hand the formula for the abdomen of the male 

 of Phyllodromia, for example, would be : — 



+ + - 

 1—7, 8 , '.), LO. 



+ + - 

 1—7, 8, 9, 10 



The sexual differences extend also to the appendages of the 

 anal piece. Thus in Periplaneta orientalis the anal valves of 

 the male are also of a transversely triangular form, but con- 

 siderably more feebly chitinized than those of the female ; in 

 the female of Phyllodromia they are similar but still more 

 strongly developed, and they bear at their lower extremity a 

 longitudinally-cleft plate, which is wanting on the soft- 

 skinned, rather globular, anal swellings of the male. The 

 sexual differences of the anal operculum {lamina supraanalis) 

 have been long since employed in classification by H. Bur- 

 meister and C. B runner von Wattenwyl. 



The movable anal appendages (cerci), as already men- 

 tioned, resemble the cephalic antenna? in their formation, only 

 they appear later and are less developed. In their structure 

 and possession of sensorial setas they also agree with the 

 antennas ; nay, from V. Graber's * experiments upon decapi- 

 tated cockroaches, their function would also seem to consist in 

 the reception of olfactory stimuli. The number of joints 

 in the cerci in the Kitchen-cockroach is 14-16, in the House- 

 cockroach 9-11. In secondarily derived forms with a more 

 globular abdomen the cerci decrease ; thus, in the Panes- 

 thidse, for example, in which the female presents only seven 

 distinct dorsal and ventral plates, they appear only as short, 

 inarticulate, triangular appendages. The late development 

 and frequent reduction of the cerci seem to show that they 

 are old inherited appendages which are approaching abortion 

 (through disuse). 



On the ninth ventral plate of all embryos and young 

 animals of both the House- and the Kitchen-cockroach short, 

 rigidly setose, unjointed appendages are to be seen which 

 distinctly originate from the ninth segment |. In the female 

 young forms of Periplaneta which are still destitute of rudi- 

 ments of wings these styles may be detected even after the 

 retractation of the last ventral plates ; they are seated upon 



* Biol. Centralbl. Bd. v. p. 452. 



t Cholodkovsky (/. c. p. 04) ascribes their origin to the tenth segment 

 and supposes them to become converted into the genital hooks of the 

 nmle ; both these views are founded upon errors of observation. 



