92 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxvi. 



This fact accounts largely for the man}' changes which have been made 

 in the classification of this group. It is now generally admitted that 

 their action is largel}' suctorial. They exhibit structures which seem 

 to show a transition from a mandibulate to a haustellate form, and for 

 this reason are of peculiar interest. 



As a whole the mouth apparatus appears as a broad, unjointed cone 

 attached to the extreme posterior edge of the under side of the head, 

 being carried so far back that its attachment to the rest of the head 

 lies largely under the pronotum (Plate X, fig. 111). The apex of the 

 cone is usually quite sharp, but never as slender as in the Heraiptera, 

 and lies, when at rest, in a depression of the prosternum between the 

 fore cox*. In many species the mouth cone is bluntly rounded. In 

 the Terebrantia it is attached to the frons by a strongly chitinized 

 thickening, running more or less obliquely across the under side of the 

 head. In most species this dark thickening is nearer the left eye than 

 the right and is connected by a similar thickening with the margin of 

 the left eye (Plate XI, fig. 120). This connection is wanting on the 

 right side, though a portion of the thickening still remains close to the 

 right eye. In theTubulifera the base of the mouth cone is much more 

 nearly symmetrical and the connections with the eyes are entirely 

 wanting (Plate XI, fig. 127). 



Axi/inmetry. — So far as we can learn, Prof. H. Garman was the first 

 to call attention to the very peculiar asymmetry which is characteristic 

 of the mouth parts of the members of this order, and he gave a new 

 interpretation to certain of these parts, which we believe to be correct. 



Not only is the connection of the mouth cone, as a whole, with the 

 frons asymmetrical, but also some of the individual parts of the mouth 

 are markedlj^ so. The most striking of these are the form of the 

 labrum and the absence of the right mandible. These parts will be 

 considered more in detail by themselves. 



Lahrum. — The labrum forms the front wall of the cone (Plate XI, 

 figs. 120, 127). It is decidedly asymmetrical in all Thysanoptera, but 

 especially so in the Terebrantia. It is irregularly triangular in form, 

 does not reach to the endocranial thickening, but is attached by its 

 broad base to the clypeus by an indistinct membranous connection. 

 From the base it narrows to the tip, where it is more or less rounded 

 in Terebrantia, but is quite pointed and spine-like in many Tubulifera, 

 though l)luntly rounded in others. It is drawn out much farther 

 toward the right cheek than toward the left, and on the right side also 

 approaches most closely to the transverse thickening. The labrum is 

 usually abruptly darker in color than the area between its base and 

 the transverse thickening. 



Ma.v'dlx. — The mixillte are broad, flat, and external. Like the 

 labrum, they are wedge-shaped or triangular in general form, and they 

 constitute the side walls of the mouth cone. They taper toward their 



