66 Mr. D. Sharp on the 



method of numbering the tribes, groups, &c., instead of 

 naming them. 



I had intended making a comparison of Dr. Horn's 

 tribes with those used by C. J. Thomson and Duval, but 

 this I find to be impossible owing to the fact that the 

 tribes of these authors consist of European insects, 

 while Horn's are made chiefly on the North American 

 fauna ; moreover, as I have already stated, the same 

 names do duty for very different values ; thus these 

 two facts render any positive comparison almost im- 

 possible. 



In considering the families composing the carnivorous 

 series. Dr. Horn adopts no less than seven families, 

 Cicindelidce, Carahidce, Haliplidce, Amphizoidce, Pelo- 

 biidce, Dytiscidce, and Gyriiiidce. The first reflection 

 about these families which occurs to one acquainted with 

 the subject, is that they are of extremely different 

 values ; thus the word CidndeUdce represents about 1000 

 species and forty genera ; the word Carahidce about 

 10,000 species of 600 genera ; the word Haliplidce about 

 fifty sijecies of three genera ; Amphizoidce two or three 

 very closely allied species ; Pelohiidce two or three 

 species ; Dijtiscidce about 1200 species arranged in eighty 

 genera ; and Gijnnidce 150 species of seven or eight 

 genera. Here again Dr. Horn follows the system in 

 vogue, and it is that system we must blame if we con- 

 sider that this process of making single species appear 

 the zoological equivalent of ten or twenty thousand 

 species is an erroneous one. There is, however, an 

 important fact indicated by this nomenclature, viz., that 

 certain species are as distinct in their structure from the 

 great complex masses of species, as these latter are from 

 one another ; in this sense we see that the isolation as 

 regards intermediate forms of a single species may be as 

 great as that of ten thousand species from another ten 

 thousand species. This is a fact of great importance, and 

 it is clear that a method of nomenclature and classifi- 

 cation that reveals instead of concealing these facts is a 

 great improvement. A jjurely synthetical classification 

 gives us these advantages ; under it Anipliizoa insolcns 

 would not be "a family," but would remain in classifi- 

 cation an isolated creature until the synthesis of family 

 value were reached. Putting aside, however, this 

 technical objection, for which, as I have said already. 

 Dr. Horn is in no sense exclusively responsible, we ma}^ 



