EXTENSIVE GENERA FAVOURABLE FOR STUDY. 219 



for so the wide intervals between the objects they 

 contain proclaim them to be, — may be set aside for 

 future analysis, when we have attained to higher 

 degrees of inductive generalisation. It is those 

 genera which, from containing numerous species 

 and modifications of form, are usually termed per- 

 fect, which we are more especially to select as fit 

 objects for the preceding line of enquiry. We all 

 know, that the more numerous and varied are the 

 materials given to us for accomplishing a given 

 work, the greater will be the degree of accuracy 

 attending the result, provided we use these materials 

 for the purposes for which they were designed, and 

 make each fit into the other with symmetry and 

 order, so as to produce a perfect whole. Applying 

 this to the question before us, we may safely assume 

 that extensive genera are the most calculated to 

 elicit the first principles of classification. They 

 seem as if intended for natural storehouses, wherein 

 we should find all sorts of implements with which 

 we may try our hand at combining, changing, and 

 remodelling, until we make all the parts, like those 

 of a complicated puzzle, fit into each other; and 

 produce, from what appeared a heterogeneous as- 

 semblage of isolated objects, a perfect tablet of 

 order and beauty. It must be confessed, indeed, 

 that such genera are the plague and torment of 

 those who seek only to arrange them artificially; 

 because the interchange of characters is so gradual, 

 and the intervals between the more prominent types 

 so filled up and crowded by connecting species, that 

 it seems utterly impossible where to make one 

 division begin, or where end. It is clear, however, 



