xxii INTRODUCTION. 



istic, as for instance the spots on the scutellum and of the post- 

 scutellum. Thus : — 



Range together all the species with post-scutellum alone yellow. 

 " " " scutellum alone spotted yel- 



low. 



" " " " " scutellum and post-scutellum 



both spotted, etc. 



Then come the anterior and posterior borders of the prothorax; 

 then the stripes of the abdomen, the spots of the metathorax, etc. 



Thus you come i<> form more species than are really existing, 

 every variety becoming a species ; but, no matter, the work of 

 reducing will at last arrive and will be easily carried through 

 when species shall have been studied and known, as it were, by 

 heart. 



Then we may look into books and try to determine each species 

 while having, beneath one's eyes, the ensemble of individuals. 



Bere we can get on only by groping about a long time in 

 uncertainty. Often after comparing these similar individuals we 

 shall detect two species which at first had been confounded. 



The determination can be carried on by two inverse proceed- 

 ings. After having sorted out apart by one's own observations 

 the greater number of species, the entomologist often knows them 

 already sufficiently to guess to what species such and such a 

 description belongs, He can, therefore, either start from a species 

 ami seek the descripl ion which suits it, or start from a description, 

 ami in passing in review his series seek to what species this de- 

 Bcription is adapted. 1 .Moreover, these two manners of proceeding 

 e into each oilier repeatedly, when one has beneath one's 

 eyes, a large ensemble of well-separated species. Indeed, when 

 making use of the first proceeding, one often falls into the second, 

 for one is often struck with the description of such and such a 

 -alient character which one recollects having observed in a species. 

 Thus the search after one species leads to the determination of 

 another. The work hastens along with accelerated motion and 

 leads to results the more certain as each species is represented 

 by a larger number of individuals comprising most of the possible 



1 For tin's purpose short descriptions, i. e., diagnoses, which may at 

 once he OOmmitted to memory, are very useful. 



