VERIFICATION OF GROUPS. 289 



tribes and orders of the whole class, we bring forth new 

 and unexpected proofs of the harmonious simplicity of 

 nature, and demonstrate our group by a mass of evi- 

 dence perfectly unanswerable; but, to establish analogies, 

 we must not be left to vague suppositions or fanciful 

 conceits, and this obliges us to test our group by de- 

 termining its types. 



(351.) III. The system of representation, by which 

 the types or divisions of a natural group are determined, 

 is the third and last test. It is by this that we can judge, 

 whether our group is perfect or imperfect ; and by this 

 we can calculate, from analogy, the probable extent of 

 the gaps that may occur in a natural series. It is easy 

 to divide the smallest circular groups, into three, four, 

 five, or seven divisions, for the propriety of these 

 divisions (they not being circular) cannot be con- 

 troverted by the answer that has been given, of "putting 

 them to the test of returning into themselves." The 

 naturalist, therefore, requires something more to guide 

 him in correctly dividing his group, not only as to the 

 number of its divisions, but as to the structure or pecu- 

 liarities which each should possess. Without this guide, 

 he will be unable, in many cases, to establish the parallel 

 analogies, and he must wander in all the uncertainties 

 of conjecture. Hence it becomes necessary to compare 

 his divisions with the characters of the types in the 

 animal kingdom. If these exhibit a conformity, how- 

 ever remote, whether in their structure, nature, or habits, 

 and if they follow each other in the same progression, he 

 has no need of looking farther. His group is one of 

 Nature's ; discovered, it may be, by himself, but existing 

 " from the beginning." 



(352.) But theory without analysis is like precept 

 without example : we shall choose, therefore, an illustra- 

 tion to show the full force of these remarks. For this pur- 

 pose we select one of the best known of our native birds, 

 the hedge-sparrow (Accentor modularis Cuv.) ; chiefly 

 because we have felt much interested in the delightful 

 history which has been given of it by the amiable and 

 u 



