320 NATURAL ARRANGEMKNT OF INSECTS. 



where the antennae are remarkably developed, and the 

 head completely hidden under the dilated margin of the 

 thorax, is probably the most typical form : it contains, 

 no doubt, its own sub-genera ; but the variations of the 

 antennae are so numerous, differing in almost every spe- 

 cies, that no reliance must be placed upon these mo- 

 difications, for sub-generic characters.* It is generally 

 imagined, that the great majority of the typical Lampy- 

 ridre, like those of the South of Europe, are luminous : 

 this supposition, however, is against our personal expe- 

 rience. We have seen, on a summer's night, in the 

 suburbs of Genoa, the air perfectly illuminated with 

 millions of one of these species; but this we never wit- 

 nessed in Brazil. We do not remember, indeed, to have 

 remarked the luminosity of any one of the numerous 

 Lampyridce found in that country, during any part of our 

 travels in Tropical America. Several other forms, more 

 or less related to these insects, enter within the family : 

 those, for instance, having theexserted head of Latreille's 

 Pyrochroides, — as our Pyrochroa riibens, where the tho- 

 rax is still nearly orbicular, but reduced to the ordinary 

 dimensions, — serves to indicate the passage between this 

 family and the CantharidcB ; while the approach, on the 

 other side, to Lycus, is still more evident. Latieille, 

 indeed, includes the latter insects as a part of the Lam- 

 pyridcB; and the analogies of both may, possibly, show he 

 is quite correct in this, — particularly as we should then 

 have but little scruple in raising his Cleri to the rank of 

 one of the aberrant families : but at present we are 

 chiefly guided by synthesis ; and when we see so strong 

 a resemblance between many of the Brazilian species of 

 Lycus and of Hispa, we cannot but suspect that these 

 two groups, standing at the confines of their respective 

 circles, connect the Monilicornes with the Malaco- 

 dermes, — in other words, that this is a relation of affi- 

 nity, rather than of analogy. 



(287-) The next typical family, or the Cantharidce, 



* See figures 4, 5, and 6. on plate 39. of Griftith's Cuvier, vol. xiv. 



