218 REPOET — 1844. 



function. More definite principles of classification maj' hereafter be dis- 

 covered, and meantime all that we can do is to arrange our systems accord- 

 ing to sound reason and without theoretical prepossession. By care and 

 judgement much may be done to give greater regularity and exactness to our 

 methods of classification, either by introducing new groups where the im- 

 portance of certain characters requires it, or by rejecting such as have been 

 proposed by others on insufficient grounds. At the present day many authors 

 are in the habit of founding what they term " neio genera " upon the most 

 trifling characters, and thus drowning knowledge beneath a deluge of names. 

 As this is a point of great importance to the welfare of zoology in general, 

 I may be excused for dwelling on it for a few moments. 



In the subdividing of larger groups into genera, even in the strictest con- 

 formity with the natural method, there is evidently no other rule but conve- 

 nience to determine how far this process shall be carried. However closely 

 the species of a group may be allied, yet as long as anj^ one or more of them 

 possess a character which is Avanting to the remainder, it will always be in 

 the power of any person to partition off" such species and to give them a ge- 

 neric name. Take the very natural group Parus for instance, as restricted 

 by most modern authors (i. e. Parus of Linnaeus, deducting JEgithalus 

 and Panuriis). First we may separate the long-tailed species, and follow 

 Leacli in calling it generically 3Iecistura. Of the remaining Pari, we may 

 make a genus of the crested species (P. cristatus), then another of the blue 

 species with short beaks (P. cceruleus, &c.), a third of the black and yellow 

 group (P. major, &c.), and a fourth of the gray species (P. palustris, &c.). 

 (JN.B. Generic names have actually been given to these groups by Kaup in 

 his ' Skizzirte Entwickelungsgeschichte der Europaischen Thierwelt.'] But 

 another author may go still further, and may again subdivide the groups 

 above enumerated, a process which would lead to the absurd result of making 

 as many genera as there are species, or in other words, of giving to each 

 species tioo specific names and no geneiic one. Therefore genera should not 

 be subdivided further than is practicallg convenient for the purpose of fixing 

 really important characters in the memory ; and seeing that there are already 

 more than 1000 genera provided for the 5000 species of birds (which are 

 probably all that can be said to be accurately known) it seems evidently in- 

 expedient to increase the number of genera, except in the comparatively rare 

 cases where new forms are discovered, or really important and peculiar struc- 

 tures have been overlooked. 



The precise rank in the scale of successive generalizations which shall be oc- 

 cupied by those groups which we leYva genera is then a matter oi convenience, 

 and consequently of opiiiion. Nature affords us no other test of the just limits 

 of a genus (or indeed of any other group), than the estimate of its value 

 which a competent and judicious naturalist may form. The boundaries of 

 genera will therefore always be liable in some degree to fluctuate, but this is 

 unavoidable, and it is a less evil than to give an unlimited license to the sub- 

 division of groups and the manufacture of names. The only remedy for 

 this excessive multiplication of genera, is for subsequent authors who think 

 such genera too trivial, not to adopt them, but to retain the old genus in 

 which they were formerly included*. 



♦ It is usual where this is done to retain the groups, which are thus deprived of a generic 

 rank, under the title o( subgenera. There appear to me however to be great objections to the 

 adoption of subgeneric names in zoology. First, it would introduce into a science already 

 overloaded by the weight of its terminology, an additional set of names «hose rank is not 

 (like that of families, subfamilies and genera) indicated by the/orinof the word, but which are 

 undistinguishable to the eye from real generic names, and would therefore be perpetually con- 

 founded with them. Secondly, subgenera would greatly interfere with the harmonious working 



,i 



