NYMPH-STAGE OF THE EMBIDiE. 377 



their enemies ; and tlie analogy is the more complete because the 

 silken galleries are not perfect tubes, but more of the nature of 

 coverings, the larva being protected above and able to feed on 

 the surface of the vegetable matters on which it rests at the same 

 time. Having in mind M. Lucas's assertions, I certainly doubted 

 if the I]mbia larvse caused any injury to the orchids; but having 

 seen pieces of roots, placed in a box with the larvae, freshly 

 gnawed, I now suspect that a mistake has been made as +o the 

 supposed carnivorous habits of the species of the group, and that 

 in reality they are vegetable feeders, as are the Termitidse ; for 

 the latter will sometimes attack growing vegetable materials. 

 There still, however, remains a difBculty in the common habit in 

 most of the species of living under stones, in which localities 

 almost the only vegetable matters that could serve them for food 

 are the mycelia of fungi or ordinary fibrous roots. 



Systematic Position, Structure, &c. 



It is possible that there is no more difficult, unsatisfactory, 

 and (I migbt add) thankless task than that of having to suggest 

 the sequential position of any group in the so-called Pseudo-Neu- 

 roptera. Latreille, the founder of the genus Embia, placed it as a 

 second genus in the subfamily Termitinse. But I think Westwood 

 happily seized upon its position as between the Termitidse and 

 Perlidse. Burmeister, who justly erected these insects into a 

 family (Embidae), separated the White Ants from the Stone- 

 flies by {inter alia) the highly specialized Dragonflies*. I do 

 not think the relationship between Termes and Einhia is so close 

 as has generally been accepted. Embia evidently is not sub- 

 ject to those polymorphic conditions so characteristic of Termes, 

 and wants its quadrifid labium (it is bifid in Enibia). It has 

 also much of the external form of the Perlidse, especially of the 

 genus Leuctra (which it resembles also in its extreme agility); 

 and external form is not always to be disregarded in searching for 

 affinities. But there are other and wide discrepancies ; and the 

 only object of these remarks is to suggest that Termes and Emhia 

 have less in common than is generally supposed. 



* It is scarcely worth while to seriously notice Eambur's extraordinary ar- 

 rangement. The Embidae follow the Termitidje ; but the Perlidag, with utter 

 disregard of all aiSnities (excepting aquatic habits), are placed between the 

 Sialidge and Phryganidas, with all the true Neuroptera Planipennia intervening 

 between Termes and Sialis. 



