220 Mr. n. S. Bagnall on the Classification 



23i Gory dor as eques. 

 Steind. SitzungslD. Akad. Wien, Ixxiv. 1877, p. 140, pi. xii. fig. 3, 



Deptli of body 2^ in tlie length, length of head 3^. Diam- 

 eter of eye 4 in length of head, interorbital width 2^. Snout 

 as long as postorbital part of head ; suborbital very deep, 

 reaching upper lip j barbel reachhig gill-opening. Dorsal I 7 ; 

 spine nearly as long as head ; free edge of fin convex, with 2 

 rays longer than spine 5 3 median scutes before adipose fin. 

 Anal I 7. Caudal emarginate. Pectoral spine reaching 

 middle of pelvics. Scutes ^^^ ; humeral shields large, 

 reaching base of pelvics and meeting below. Brownish 

 above, yellowish below, with a broad blackish lateral band 

 tapering backwards ; fins immaculate. 



Amazons (Teffd and Cudajas). 



XXV. — Some Considerations in regard to the Classification 

 of the Order Thysanoptera. By RiCHARD S. BagNALL, 

 i\L.S., F.E.S., Hope Department of Zoology, University 

 Museum, Oxford. 



Since my papers on the Urothripida? were published *, I 

 have come to the conclusion tliat in retaining that family as 

 a family of the suborder Tubulifera unnecessary difficulties 

 will be created. I have already shown that whereas Uro- 

 thrips superficially resembles the Tubulifera very closely, it 

 really differs from true Tubulifera more strongly in its 

 structure than do the members of the suborder Terebrantia ; 

 or, in other words, the two suborders Tubulifera and Tere- 

 brantia are more closely related to each other than Urothrips 

 to either. 1 am now convinced that the only course one can 

 reasonably take is to erect a new suborder for the reception 

 of the family Urotlnipidse, for which I propose the name 



POLYSTIGMATA, 



suggested by the character that appears to me to be of the 

 greatest taxonomic value. 



It will be well to briefly diagnose the three suborders. 



« Ar.nales Musei Natioiialis Hungarici, 1909, vii. pp. 125-136, pi. iii., 

 it Mem. 1" Congres Intcmalioiial d'Eiitoniologie, 1910, pp. 283-288. 



