NO. 1310. NORTH A MERICAN THYSANOPTERA— HINDS. 219 



ida3, while in PhlcBothripidjvi it is entirel}' wanting. Moreover, there 

 exists in Thripidte a wide variation in the degree of its development, 

 as has alread}^ been shown. So in this respect, also, we must place our 

 three families in the same relation to each other, and if Prothysanop- 

 teron possessed an ovipositor, as we can not doubt from its ph yllogeny 

 must have been the case, the well-developed organ found in ^'Eolo- 

 thripidiv must very proljabl}^ approach most clos(dy to tlic primitive 

 form. 



Summarizing the conclusions which we have now reached, we tind, 

 first, that the Tubulifera (Phheothripidie) have diverged more widely 

 from Prothysanopteron than have either of the families of the Tere- 

 brantia. Second, a comparative consideration of antenna^ wings, and 

 ovipositor shows that ^l^]olothripid{« and Phkeothripida? present the 

 extreme types of these structures found in the order. Therefore we 

 conclude that the ^^olothripidie most nearly' preserve the characters 

 present in the Prothysanopteron ancestor of this order. From this it 

 appears that the descendants of Prothysanopteron early divided into 

 two main groups, one of which diverged widely from the original form 

 and has developed the Tubulifera of to-day. The other of these 

 groups continued nearly along the original line, but in time it divided 

 again and a group (Thripidte) branched otf, taking in some respects the 

 direction of Phloeothripidiv, while in the majority of characters it fol- 

 lowed a line of its own. The group which still continued most nearly 

 in the original direction includes the insects which we now place in 

 the family ^T^Jolothripida?. 



