6 INTRODUCTION. 



man ought to possess, require that we should not allow 

 ourselves to be dazzled by the seeming simplicity of rules, 

 so far as to overlook the cases where these rules interfere 

 with the evident order of nature. I am induced the rather 

 to make this observation, because I have long felt that the 

 system adopted by Geoffrey, Dumeril and Latreille, of 

 arranging insects according to the number of joints in 

 their tarsi — a system now in very general use, owing to 

 the unrivalled reputation of the last-mentioned entomo- 

 logist — is by no means natural a . By the inspection of the 

 foregoing table it will appear, that few characters are so 

 variable as those drawn from the tarsi, and consequently 

 few so objectionable for general use ; though it must be 

 granted, that in particular tribes these organs may safely 

 be used for generic distribution. The consequence, how- 

 ever, of making primary divisions after the number of 

 joints in the tarsi, has been, that many genera have been 

 separated and thrown at an immense distance from the 

 true place assigned them in nature. Among numberless 

 general instances of which evil it will only be necessary 

 to note the affinity between the Psephalidcs, Leach, and 

 Stapliylinida, Kirby, the genera Cis, Lat. and Anobium^ . y 



a " Ce n'est pas assez pour pallier les frequentes anomalies qui se ren- 

 contrent, de dire que tel insecte a cinq articles aux tarses quoiqu'on n'en 

 decouvre que quatre. 11 devroit en effet en avoir cinq d'apres la regie 

 qu'on a etablie. Cependant ces explications, toutes ingenieuses qu'elles 

 sont, ne peuvent me convaincre, et je ne puis me determiner a supposer 

 plus d'articles aux tarses qu'on ne peut reellement en decouvrir, en ymet- 

 tant toute l'attention possible. Le nombre des articles doit etre clair, 

 positif et uniforme, sans quoi le caractere est vacillant. Au reste, je ne 

 disconvietidrai pas que sans etre rigoureusement propres h diviser les 

 ordres, ils ne puissent tres-bien servir pour un tableau dans lequel les 

 genres seroient ordonnes selon le nombre des articles des tarses." — Enh 

 Helvetique, p. 30. 



