NO. 2242. MORPHOLOGY OF THE 8TREPS1PTERA— PIERCE. 407 



The mandibles show some fair characters in the genus Stylops 

 and are used in the tables. The principal variation is in the num- 

 ber and position of the teeth. (See pi. 71.) 



The spiracles furnish good characters as to number, position, and 

 form. The Callipharixenidae have two pair of spiracles on the 

 cephalothorax (pi. 68, figs. 1, 2). The Xenoidea have otherwise only 

 one pair, and it is presumable therefore that this family are not 

 Xenoid but rather Mengeoid. Abdominal spiracles have not been 

 noticed except on Dacyrtocara undata^ which has three pair. (pi. 

 74, figs. 5, G.) 



The median unpaired genital tubes are so often impossible to find 

 because of the condition of the material that it is the writer's practice 

 to note them whenever observed. In addition to the records already 

 published it may be noted that CalUpharixenos muiri, Chrysoco- 

 O'ixenos siaTneTisis, and /Stylops vicinae have five tubes, and Dacyrto- 

 cara undata has two (pi. 74, figs. 5, 6). 



In the Dacyrtocara the first abdominal segment extends far in 

 front of the cephalothorax, but within the host's abdomen (pi. 74, 

 figs. 5, 6). 



The female type of Chrysocorixenos siaw^nsis has an asymmetrical 

 cephalothorax, there being a sort of tumor near the base of one side. 



The writer has from the beginning of his work on this order at- 

 tempted to find other characters beside those of the appendages for 

 use in classification. Although it was quite apparent that the legs, 

 antennae, mouth parts, and genitalia gave a satisfactory and logical 

 system of classification, there was always the possibility of not 

 being able to identify the insect if the appendages were lost. 



The dorsal thoracic characters were delineated in the specific de- 

 scriptions in Bulletin 66, and transferred to the generic descrip- 

 tions in Genera Insectorum. But the great divergence of thoracic 

 structure did not seem to permit their use in family descriptions or 

 phylogenetic studies. 



Since the last contribution on the order many new species have 

 been received and a study of these with a review of material already 

 described now makes it possible to give a clearer treatment of the 

 comparative morphology of the group. 



Progression of characters. — Certain striking trends of modifica- 

 tion are apparent. The antennae may be considered to have had 

 typically seven joints with at least the third laterally produced, 

 flabellate. These appendages are found in the order with any num- 

 ber of joints from four to seven, and with from one to five joints 

 flabellately produced. 



