14'4 MODERN CLASSIFICATION OF INSECTS. 



tylus.* If, however, the Erotyhdae be thus introduced, the Endomy- 

 chidae appear to me to be too closely allied to them to be rejected. 

 Mr. MacLeay, indeed, suggests that it is by the Erotylidse that the 

 opposite points of the circle of Coleoptera meet. One thing is, 

 however, certain as regards these insects; namely, that although the 

 tarsi in the Coleoptera generally offer a decided constancy of cha- 

 racter in exclusive groups, they here exhibit numberless variations, 

 even in the sexes of the same species (^Jig- 12. 9. 10. and 17. 18.). 

 From these considerations, and from the very variable structure of 

 the insects introduced into the family Engida>, especially in their 

 larva state, it will be evident that the family requires much subdivision ; 

 a step which ought cautiously to be taken. With a view to effecting 

 this, I have dissected many of the genera, and have been fortunate in 

 procuring a considerable number of their larvae ; but I would not wish 

 the groups subsequently given to be regarded otherwise than as pro- 

 visional. 



We have already seen that some of the genera, such as Ips and its 

 allies, are nearly allied to the Nitidulidae. The general habits of these 

 and other approximating genera indicate a group, of which, both as 

 regards the larvae and form of the trophi, I have regarded Trogosita 

 to be the type. In the majority of these the maxillae are terminated 

 by a large single lobe (yfi(). H. 17. and 12. 4.), and the larvae are 

 more or less fleshy, with two short curved hooks at the extremity of 

 the body. The affinity of these insects with the Cucujidae, and of the 

 latter with the Paussidae, I have endeavoured to prove in the Zool. 

 Journ. and Linn. Trans, There still, however, remain a number of 

 genera which I would provisionally arrange with Mycetophagus, al- 

 though I cannot but admit that they have but little real relation there- 

 with. The Erotyhdae, notwithstanding their evident relation with 

 Engis and Dacne, I have provisionally removed from the family. 



As regards the habits of these groups, it may be generally observed 

 that they are found under the bark of trees, or in fungi, chiefly feed- 

 ing upon decaying vegetable matter; never attacking either living or 

 dead animal matter. 



* See the observations of M. Lacordaire, upon the strong affinity existing be- 

 tween Engis and Erotylus, both in form and habits. In the larva state, the former 

 reside in Boleti, &c., and it is also upon or in the neighbourhood of such pro- 

 ductions that the perfect insects are found in South America. (^Ann. Soc. Ent. 

 France, vol. i. p. 359.) 



