Classification in Zoology. . 227 



body, and not the stove or grate. How is it possible for a grate 

 containing burning coals to be insulated, so as to retain a charge of 

 electricity 1 On the other hand, it is presumed that the experi- 

 menter was insulated by standing upon a carpet made quite dry by 

 a winter fire. — American Journal of Science and Arts, Vol. x., 

 No. 30, p. 321. 



On (he Principles of Classification in Zoology. By Professor 

 L. Agassiz. 



It may be said that investigations upon the structure of 

 animals have already yielded all the information coming from 

 this source which can serve to improve our classification of 

 the animal kingdom. 



After the great general divisions of the animal kingdom 

 have been circumscribed in accordance with their anatomical 

 structure, after the classes of the animal kingdom have been 

 characterized by organic differences, it is hardly possible to 

 expect that further investigations upon the structure of ani- 

 mals will afford the means of establishing correctly the 

 natural relations of the families. For it is already seen that 

 the amount of organic difference which exists between the 

 different families is either too insignificant to afford a test by 

 which to settle their pre-eminence or inferiority, or so striking 

 as to impress us with an exaggerated idea of their difference. 

 Many examples could be quoted to shew, that, in this respect, 

 from the same identical facts, naturalists have arrived at 

 very opposite conclusions. And this diversity of opinion 

 among investigators of equal ability leads me to think that 

 comparative anatomy has done its work in that direction, 

 and that we must seek for another principle in order to settle, 

 in a natural way, the respective positions of the minor divi- 

 sions throughout the animal kingdom, and to set aside, once 

 for ever, the arbitrary decisions which we are constantly 

 tempted to introduce into our classifications, whenever we 

 attempt to arrange all the families in natural groups. Be- 

 fore so much had been done to improve the natural classifi- 

 cations of the animal kingdom, it was hardly possible to 

 notice how much was, on every occasion, settled by induction, 



p2 



