NO. 1310. NORTH AMERICAN THYSANOPTERA— HINDS. 215 



powerful development of the muscles of the fore leo^s or of any modifi- 

 ciition of the tarsus which would tend to give a iirmer hold in crawling, 

 being favorable to the insect, would be preserved ])y natural selection, 

 and thus in the course of many generations the tarsal tooth and the 

 powerful, thickened femora of most Tulndifera would be developed. 

 There would also ])e a cori'elative broadening and flattening of the 

 prothorax, which would nc^cessarily result in pushing farther apart the 

 fore coxa^, which are attached to its hind angles. The logical result of 

 these changes is the trapezoidal form of the prothorax always found 

 in those species having such thickened femora and well-developed 

 tarsal hooks. 



The elongation and flattening of the body are doubtless referable in 

 soiue degree to the same change in the conditions of external life, for 

 such a changed form would certainh^ have been favorable to its pos- 

 sessors, and we are surely .safe in assuming that the favorable changes 

 are the ones which have been preserved, while the unfavorable ones 

 have been eliminated. We do not presume to say that all the descend- 

 ants of Prothysanopteron followed this suggested line of change; some 

 of them certainly may have done so. Neither do we presume that all 

 the descendants of those which did follow some such line of develop- 

 ment would continue in an even similar environment till all the modi- 

 flcations which have l)een named had been accomplished. We have 

 just as nnich reason to expect a change of environment anywhere along 

 the phyllogenetic line as at its beginning, and such changes certainly 

 must have taken place. What would be the result if this were the 

 case? Difl'erent enviroiuuents acting upon different subjects, or even 

 upon like sul)jects, w^ould favor entirely different variations. Struc- 

 tur(>s which had become dexeloped during the changes subsequent to 

 Prothysanopteron might l)e lost, but those that had been lost could 

 never again be developed in their original form; e. g., tarsal teeth 

 and thickened femora might develop and then disappear, but an ovi- 

 positor of the original type would never again be found in the Tubu- 

 lifera. We would expect then that the descendants of Protubuliferan 

 would vary in habits, habitat, form, and life rather than in the tubu- 

 lar nature of the terminal segment of the body. Such is indeed the 

 case, and so while there do take place great modifications of each 

 organ, the presence of the tube is constant. We feel justified in con- 

 cluding that the family Phhjeothripida^ has now diverged far more 

 widely from Prothysanopteron than has either of the families of the 

 Terebrantia. 



The two families constituting the suborder Terebrantia resemble 

 each other (juite closely in many respects. We find between them no 

 such marked points of difference as we do between each of them ancJ 

 the Phlwothi'ipidje. The principal ditferences which do exist ar/e 

 mainly various moditications of the .same organ, and the most impor- 



