156 MR. E. m'lachlan on the neueopteea or 



and IscTinura senegalensis*) are African species not kaovrn to 

 occur in Europe. 



Thus we find about 25 species that, so far as our present 

 knowledge serves, are peculiar to tlie islands. It is quite pos- 

 sible that before the destruction of the forests, wlien the islands 

 were better watei-ed, the number of species was greater ; it is 

 also certain that much yet remains to be discovered. For instance, 

 it is difficult to believe that the smaller forms of Perlidse (Leuotra 

 and Nemourd) are totally absent ; so also it is difficult to believe 

 that no Termitidee exist in the Canary Islands; and several other 

 examples might be cited. 



That some of the purely terrestrial forms may have been intro- 

 duced from Europe is ver'y possible ; on the other hand, I see 

 no reason to doubt that some of the European forms mentioned 

 may be considered true natives o£ the islands also ; and it is still 

 further possible that some forms apparently peculiar to the islands 

 may yet be discovered in Southern Europe or North- western 

 Africa ; for in some respects we know less of the Neuropterous 

 fauna of these regions than we now do of the islands. It is 

 worthy of remark, however, that (with one possible exception) 

 the whole of the species of Trichoptera are peculiar to the islands 

 although belonging to familiar European genera, and that they all 

 inhabit running water in the larval stage. These remarks show 

 that it would be hasty to attempt generalizations from the Neu- 

 ropterous fauna alone. With the Coleoptera it is different ; and 

 those who wish may obtain generalizations on these, from two 

 very opposite points of view, by consulting Mr. Wollaston's ' Co- 

 leoptera Atlantidum ' (and his other works), and Mr. Wallace's 

 ' Geographical Distribution of Animals.' 



TRICHOPTEEA. 



LlMKOPHILID^. 



LiMNOPHiLus APPiNis, Gurtis. {L. cinctus, Hag. Eut. Month. 

 Mag. ii. p. 75 ; Stett. ent. Zeit. 1865, p. 217.— i. f#7n>, M'Lach. 

 Eevision & Synopsis, p. 82.) 



Madeira {Wollastoii). 



Renewed examination has not revealed any tangible differences 



between the Madeiran examples (Z. cinctus) and ordinary L. affi- 



nis. There is a certain amount of not-describable colour-variation 



* To these it is possible that Libellula chrysostigma, Burm., should be added. 



Compare the detailed i-emarks on that species at pp. 177-179, post. 



