CERAMBYCIDAE. 269 



Several cliaracters which have been recently adopted for the 

 difFerentiation-of tribes seem to be of but small, or still worse, 

 illusory importance; and among these, the extension outwards 

 of the middle coxa?, so that they attain or not the cpisterna, is 

 one of the most indelinite, and we have, therefore, rejected it as 

 far as possible in the following scheme. 



We have, in common with previous investigators, failed thus 

 far to find any distinct difference capable of expression in words 

 between this family and Chrysomelidae. One familiar with the 

 subject will rarely if ever mistake one for the other. But so far 

 the essential difference between the Tetramera, of which the hirva3 

 feed upon wood, and those feeding upon cellular vegetable tissues 

 has eluded observation. We can merely at present observe that 

 a slight approximation to it seems to be made in the fact, that 

 in the Cerambycidte there is a tendency in the epimera of the 

 metathorax to extend to the sides of the ventral segments, while 

 in the Chrysomelidae the 1st venti'al is prolonged forwards at the 

 sides to meet the metathorax ; thus showing probably a lower, 

 though necessarily more recent, type, which could have existed 

 only since the development of the higher broad-leaved plants. 



And in continuation of this same subject, we would refer the 

 difficulties of classification of the Longicorns to the fact, that 

 being exclusively feeders upon woody tissue, and passing a very 

 long period in the larval state, in the interior of trunks or 

 branches of trees, protected against inundations by the buoyancy 

 of their juvenile homes, they have been peculiarly qualified, not 

 only for an early introduction, but prolonged existence; and that 

 we, therefore, have here a more perfect record than is likely to 

 occur in any other land animals. Among marine objects frequent 

 examples occur of the representation in the existing fauna of 

 forms more fully represented in previous geologic ]ieriods; but 

 this is the first instance in which we have had occasion to note 

 the probability of its occurrence in the Coleoptera. Dr. IjeConte 

 has already alluded to this subject,* especially in connection with 

 the SpondylidfB, and we are very glad to find that the idea has 

 been approved of by our friend H. W. Bates,f the distinguished 



* An attempt to Classify, etc., Journ. Acad. Nat. Sci. 2(1, ii. 00 (IS'.l). 



t Contributions to an Insect Fauna of the Amazon Valley, Coleoptera, 

 Longicornes, Part I. Lamiaires, p. 5-G (from Annals and Mag. Nat. Hist. 

 18(51). 



