of the Family of Anatida. 13 



object to these arbitrary divisions. If this error led to so much 

 confusion'and to such forced combinations before any pretence 

 was made to discover the natural system, it is fraught with 

 peculiar evil in these days, when a glimpse of the harmonious 

 plan of the GREAT CREATOR is the ultimate object of every 

 zoologist. By fixing a priori upon a type, and arranging all 

 the subordinate groups in their several supposed relations to 

 that type, we not only incur the risk of commencing in error, 

 but of continuing it through the whole contents of a very large 

 division, and an entire order of animals will thus be inevi- 

 tably thrown into confusion. 



It must indeed be admitted that, with our present imperfect 

 knowledge of the laws of nature, we must, in many instances, 

 be content with theoretical deductions: but we should ever 

 bear in mind, that these theories must be relinquished when 

 opposed to a better knowledge of facts, and the right interpre- 

 tation of such facts. It has been well observed, that every great 

 discovery has originated from some preconceived theory which 

 has struck open a new path of inquiry, enlarging and expand- 

 ing the mind at every step. No one will, therefore, deny the 

 importance, much less the legitimate use of theory, as an instru- 

 ment necessary to develope the greatest truths : the objections 

 are not to its use, but to its abuse. Every naturalist, in fact, 

 who does not confine his attention to the sole study of species 

 must have a theory or a system : it may be borrowed, or it may 

 be new, true or false ; still, so soon as he attempts combina- 

 tions, he becomes essentially a theorist. He sorts and sepa- 

 rates his species into parcels or groups, under some definite 

 notions as to the characters they possess. If, after having 

 done this, he proceeds to verify his first impressions by criti- 

 cally examining their correctness, or, in other words, by ana- 

 lyzing his groups, he makes that just and legitimate use of 

 theory which is allowable : but, on the other hand, if he con- 

 tents himself with investigating the details of one of his groups, 

 and, having discovered certain properties belonging to it, pro- 

 ceeds to apply these results to the remainder, taking it for 

 granted that further analysis is unnecessary, he (hen invests 

 theory with an importance to which it is not entitled. So 



