Classified tion , 1 949 



again take place. An interesting and remarkable example is also 

 afforded by the marsupial series, a group united among themselves by 

 constant and well-marked peculiarities of organization, but presenting 

 differences much greater than are usually seen between members of 

 one order, but yet not sufficient to warrant the application of that 

 term as generally used to any of its subdivisions, and its close alliance 

 with the other Mammalia forbids its ranking as a class. It is not the 

 object of this essay to clear up any of the doubtful points just men- 

 tioned; but if the reader fully comprehends the meaning of the pro- 

 position, that, granting affinities and even groups to be natural, the 

 limits assigned to those degrees of difference and similarity which we 

 are wont to indicate by definite terms are not, and sees in these exam- 

 ples the slightest reason for assenting to its truth, the purpose for which 

 they are brought forward is fulfilled. 



That affinities really do exist in nature, it seems scarcely probable 

 that any one acquainted with the leading facts and generalizations of 

 Natural History would deny, although differences of opinion may ex- 

 ist as to their extent ; indeed, were it not so, — that is, were it impos- 

 sible to classify otherwise than by characters chosen in a purely arbi- 

 trary manner, — we could not expect to find even that amount of 

 agreement which actually exists among the almost countless number 

 of systems that have from time to time been offered for adoption. It 

 might on first consideration seem easy, by classing together such spe- 

 cies as present the greatest amount of resemblance, to arrange the 

 whole series according to their natural affinities, and, by mentioning 

 the points of resemblance, to characterize the different groups thus 

 formed ; but this, as every naturalist is aware, is a task of no incon- 

 siderable difficulty, for among the great number of points of resem- 

 blance and of distinction to be taken into consideration, a species may 

 be found to resemble, in different respects, a great many others, and 

 those differing considerably among themselves. It is not, then, by 

 the seizing of every character which the eye can perceive, that the 

 natural affinities are to be ascertained, since by that means the student 

 is only led into labyrinths more intricate as he advances in his re- 

 searches: on the other hand, a system founded on the peculiarities of 

 one set of organs alone, must, from its manifest incompleteness, be 

 arbitrary, and can only serve as a kind of index of reference, or as an 

 assistance to the memory. It is obvious, then, that neither of these 

 methods of classification can ever bring us to a knowledge of the na- 

 tural affinities, or to an appreciation of the characters by which they 

 may be ascertained : we must, therefore, have recourse to the study 

 v 3D 



