NORTH AMERICAN COLEOPTERA. .109 



tions will be found. So, also, with the second joint in its relation 

 to the third; in many cases the former is twice the length of the 

 latter ; often they are about equal, while others act as iutergrades. 



The relative length of the second, third and fourth joints of the 

 antennjfi is subject to considerable variation. In about half the 

 species tlie second and third joints are small, the third always a little 

 hniger than the second, the two together not as long as the fourth. 

 In two species, spretus and texanus, the second is very much smaller 

 than the third, which is nearly as long as the fourth. In the re- 

 mainder of the species the three joints are respectively gradually 

 longer, so that the second and third together are longer than the 

 fourth. 



The structure of the prosternum also requires attention. In those 

 species of the more oval form and with the widest thorax, as in 

 thoracicus and varicornis, the front coxae are absolutely contiguous 

 without trace of prosternum between them, while the more oblong 

 species have, in most cases, the coxae separated by a very narrow 

 prolongation of the prosternum. Had the length of the hind tarsal 

 joint and the form of prosternum been concurrent in all cases, 

 there might have been good reasons for dividing the genus, but, as 

 before stated, the first hind tarsal joint varies so much in its relation 

 to the other joints that nothing positive can be done with it. 



One of the minor peculiarities of the species of this genus requires 

 particular attention, and that is the coloration of the legs. It seems, 

 indeed, remarkable that such apparently trivial differences of color- 

 ation should be indicative of specific distinctness, but the persistency 

 of color, and its almost absolute invariability within specific limits, 

 is proven by the presence of other characters often sexual. 



The coloration of the antennae is less invariable than that of the 

 legs. 



The general color shows no variation within specific limits, except- 

 ing so far as the metallic-blues vary to green. 



The comparatively numerous species are from all parts of the 

 country, each species, with few exceptions, of limited distribution. 

 They may be separated by the following table :* 



* Luperus thoracicus Boh., Eug. Eesa, lus. p. 181, described as from Sao Frau- 

 cisco, has not been ideutified, aud, like very many of the species there described, 

 may not be from California at all. The name is preoccupied, and should be 

 dropped from the literature. 



TEANS. AM. ENT. SOC. XX. JUNE, 1893. 



