674 FAUNA HAWAIIENSIS 



amongst those found in England, however, we have detected a solitary example of 

 that sex. 



Vestigial Wings, etc. A distinct feature in the Hawaiian Thysanoptera is the 

 exceptionally large proportion of apterous species or species wherein the wings have 

 been reduced to a functionless pad. In the truly wingless species we find that the 

 ocelli are absent, whereas in the brachypterous forms the ocelli are seldom if ever 

 entirely lost. In some species o'i Dolerothrips the wings are reduced to such an extent 

 that it is only by careful microscopical examination that the minute bristle-set scale-like 

 pad can be distinguished, but in all species the ocelli are well-developed, larger than is 

 usual in brachypterous forms. 



Dermothrips is purely an apterous form on the islands Kauai, Maui and Hawaii, 

 but, as mentioned before, a large winged form is peculiar to a fourth island, Oahu. 



Unfortunately the material is not sufficient to warrant one in theorising on these 

 interesting questions. 



Taxonomy. Perhaps one of the greatest drawbacks in the study of the 

 Thysanoptera is the want of definition in specific and generic characters. After the 

 main divisions, which are comparatively well-marked, it has as yet been impossible to 

 lay down any plan by which the genera may be readily and naturally divided on 

 workable characters such as exist in, we think, all other orders. For instance in the 

 Coleoptera we have well-marked and invaluable means of systematic grouping in such 

 parts as mouth organs, the feet, the sternum, the abdomen, the genitalia, etc., whilst in 

 other orders these and other parts such as wing-venation, antennal characters, etc., are 

 equally important. Nor do the species of this order possess structural characters such 

 as those so beautifully exemplified in that group of primitive soft-bodied wingless 

 insects, the Collembola or Springtails, wherein we find the structure of the spring and 

 foot, the eye-spots, etc., a very valuable aid to identification. 



In fact we find several features, important in most orders, of little ta.xonomical 

 importance in the Thysanoptera excepting perhaps in diagnosing the subordinal 

 divisions. These remarks apply particularly to the sub-order Tubulifera, which so far 

 as the Hawaiian fauna is concerned affect us more closely than the Terebrantia. 



In the Terebrantia certain characters have been used in the separating of the two 

 families Aeoioihripidae a.nd Thripidac, such as the form and segmentation of the antennae 

 and the form of ovipositor in the female, but recent researches have brought to light 

 several forms possessing antagonistic characters which have somewhat weakened their 

 value. Nevertheless, compared to the Tubulifera, the insects of the sub-order Tere- 

 brantia are not so difficult to group systematically. 



The satisfactory generic grouping of the species of the sub-order Tubulifera is one 

 of much difficulty. The parts, as we have said already, that in many orders exhibit 

 invaluable characters for the satisfactory and natural grouping of the species, are in this 



