THYSANOPTERA 675 



sub-order chiefly remarkable by their similarity and want of definition. Thus the 

 sternum, the mouth parts, the feet, the antennae and the wings can only play a com- 

 paratively unimportant part in the systematic arrangement of these insects. It is true 

 that there are certain well-defined genera or groups of genera, such as Megalothrips 

 Targ.-Tozz., and allies, wherein the sixth abdominal segment is laterally produced in the 

 male; Macrol/irips Bagnall and Ecacanihotki'ips Bagnall wherein the fore-coxae are 

 curiously produced ; Diiiothrips Bagnall, remarkable for the bifurcate lateral meso- 

 thoracic appendages in the male ; Polyommaiot/irips Buffa possessing the eyes large 

 and contiguous or apparently so, and Ecacant liothrips and Eupatliithrips Bagnall each 

 having a distinctive and peculiar type of antennal sense-organ ; but the fact remains 

 that as a whole the sub-order is a difficult one to understand and classify. 



Again we find instances of two groups of species which may be separated easily by 

 the naked eye or under a comparatively low-power lens, but though one can have 

 little doubt as to the distinctness of the two so-formed genera, yet the differences are 

 exceedingly difficult to convey in words. Such is the case in the genera Idolothrips 

 and Dicaiothrips Buffa. We have had a number of species of both genera through our 

 hands from Central America, the Malay Archipelago and Africa, and whilst satisfied as 

 to the value of the genera, we have found it very difficult to formulate the common 

 characteristics of each genus. 



As to specific characters, the relative length of the head and prothorax, and of the 

 tube compared to its breadth at base and to the length of head and sometimes to the 

 length of preceding segments, are apparently good characters in most genera. In our 

 table ot the genus DicaiotJwips in a recent paper on Neotropical Tubulifera' we found 

 these characters invaluable ; and the relative lengths of the seventh and eiohth 

 abdominal segments were also helpful. 



We are, in a manner of speaking, in the early stages of this study, and it is therefore 

 essential that all authors should describe and figure each species fully, and in addition 

 briefly compare them with their allies. 



Chaetotaxy. The number, form and arrangement of bristles on the head, pro- 

 thorax and abdomen will prove to be characters of some taxonomical importance in the 

 Tubulifera as well as in the Terebrantia, and it is well to draw attention to the 

 chaetotaxy. 



As an illustration we find in the genus Dicaiothrips already referred to that the 

 post-ocular bristles are always present in the female but usually though not always 

 absent in the male ; thus we have found a useful character in our table oi the Neo- 

 tropical Dicaiothrips. 



In the material now before us we may draw attention to certain features relative to 

 the subject. Dermothrips (with one other known Phloeothripid genus) is peculiar on 



' Bagnall, Journal Linn. Soc, Zoology, xxx. pp. 369 — 387, pi. 51 — 53, 1910. 



