180 PROCEEDINGS OF THE NATIONAL MU8EVM. vol. xxvi. 



nullv. It also seems probal)le that the so-called ''rod" of the Tul)u- 

 lifeian female is but the vestig-e of a former ovipositor. 'Phe wing 

 venation also indicates that the ^Eolothripidfe come nearest the prim- 

 itive form and that Thheothripida? have diverged farthest from the- 

 type, with the I'hripidie somewhere in between. This species is 

 therefore of considerable interest as possibly being one of the guide- 

 posts to the phyllogeny of the order Thysanoptera. 



CHARACTERS OF TUBULIFERA (PLCEOTHRIPID7E). 



The members of this suborder agree so closely in general characters 

 that they have all been included in the single faudly Phloeothripida?. 

 They are, as a rule, considerably larger and more powerfully formed 

 than the Terebrantia, some of them being the giants of the order. 



In the insects belonging to this suborder the head is always as long 

 as broad, and may be two or three times as long. In most of those 

 species which have comparatively short heads the front is smoothly 

 rounded, but in those having ver}' much elongated heads the vertex is 

 considerably elevated, in some cases even forming a very prominent 

 conical projection of the vertex beyond the bases of the antennte. 

 The eyes varj^ widely in size and number of facets. Ocelli are gen- 

 erally present. The cheeks are usually nearl}^ straight and parallel, 

 and in some species set with more or less numerous spine-bearing 

 warts. Nearly every species has a pair of well-developed spines stand- 

 ing immediateh^ behind the eyes, and therefore called post-ocular 

 spines. The antennas are invariabl}^ eight segmented in the adult 

 stage and the sense cones on the intermediate segments are always 

 simple. The mouth cone varies in form, being in some species short 

 and blunt, and none of the external parts are acute at the tips; in 

 others the labrum is abruptly constricted beyond the middle, its end 

 forming a sharp spine-like process, which reaches beyond the broadly 

 rounded lal^ium; in still others the entire mouth cone, labium and 

 all, is elongated and tapers to a quite slender tip, which, how^ever, 

 is not spine-like. These different forms of mouth cone have been 

 thought to possess a generic value in classification, but my studies 

 thus far have led me to the conclusion that too high a value has 

 been placed upon this single character. The maxillary palpi have 

 always two segments, of which the basal is very short, and the labial 

 palpi are also two segmented, though frequently they are short 

 and indistinct. 



The prothorax has, in most cases, a trapezoidal form, and this is 

 especially noticeable in those species in which the fore femora are 

 much enlarged. The regularity of the outline of this trapezoid is, 

 however, more apparent than real, as will be seen by reference to Plates 

 VIII, IX, and X. The projecting fore coxae fill in the hind angles 

 so smoothly that in manj^ cases careful focusing is necessary to show 

 that the outline is not entirely that of the prothorax alone. The pro- 



