642 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxix. 



The venation of the wings of this family is like the typical wing in 

 the origin of media, the location of the medio-cubital cross- vein, and 

 the type of anal cells. 



The Xiphydriidffi has generally been considered by systematists as 

 a subfamily of the family Siricidte. That it represents a distinct line 

 of development is shown by the condition of the area between costa 

 and Sc+R+M, which is broad, and the wing membrane not any thicker 

 than that of any other part of the wing, while the remnant of snb- 

 costa retained is a modification of the type found in Xyela^ the type 

 found in the Siricidse, as will be shown later, is a modification of the 

 type found in NeuTotoma. That the Xiphydriidas represent a distinct 

 line is further shown by the origin of the base of media, by the posi- 

 tion of the radio-medial cross-vein, and by the origin of the base of 

 media in the hind wings, which is in reality quite a specialized condi- 

 tion. All these characters go to show that the wings of this family 

 are more nearly like those of the generalized Tenthredinoidea than 

 those of the specialized Tenthredinoidea. 



Siricidae,. — This family contains five genera, all of which are limited 

 to the northern hemisphere. The Siricid* are large, active, flying 

 insects, and as a result their wings are long and narrow. The wings 

 are like the typical wing (figs. 86-91) only in having homologous veins. 

 The stigma is narrow, pointed, and eight to ten times as long as broad. 

 The area between costa and Sc+R+M is narrow and almost as strongly 

 chitinized as the veins themselves. The suppression of the subcosta, 

 which is represented in some species as a pale, indistinct line, is 

 undoubtedly due to the chitinization of the membrane of this area. 

 In Tremex columha (fig. 91) the chitinization has proceeded so far that 

 there is a large trachea unprotected by a vein ramifying through this 

 area. The peculiar arrangement of the veins forming the apex of the 

 cell M has already been discussed and need not be considered here. 

 The posterior end of the radio-medial cross-vein joins the cell M^ on 

 its basal third; in Tremex fuscicoTnis it ends in the angle between the 

 transverse and longitudinal parts of media, while in Slrex californicus 

 (fig. 87) it ends on the posterior third of the transverse part of media. 

 Correlated with the migration of the posterior end of this cross-vein 

 there is a corresponding migration of its anterior end along the radial 

 sector until, in Tremex fuscicorni.s^ it arises almost at the origin of the 

 sector, so that the cell R is hardly more than a broad line. In Tremex 

 the radial cross-vein is situated near the apex of the cell Rg+R^, 

 the free part of the vein R^ is wanting. On the apex of the front 

 wings of all the species of the famil}^ there has been developed a large 

 appendiculate cell, with the apex of Rj as a prominent vein extending 

 toward the actual apex of the wing. An especially interesting feature 

 of the wings of the Siricidse is the preservation of the free part of Cu2 

 in the ^enus Paururus (fig. 86) and a j^ortion of it in difterent species 



