336 PRACTICAL AND SCIENTIFIC ZOOLOGY. 



True it is that the author in question did not think it 

 necessary to verify this group by tracing its parallel 

 analogies in the next (the Merulidce, or thrushes) which 

 succeeds to it. Had he done so, he might probably 

 have discovered that this, so far as concerns the order 

 of the divisions just specified, were in reality an arti- 

 ficial circle. And yet this conviction might not have 

 been arrived at; for, if implicit reliance were placed upon 

 the accuracy of this series, and we merely proceeded 

 to fix upon the groups analogical to these in the next 

 family circle, we should be at no loss to make them out 

 in the following manner : 



SHRIKES. Characters common to both. THRUSHES. 



larvus. 



Edolius. Feet very short. Brachypus. 



Tyrannus. Live in the vicinity of water. Crateropus. 



Ceblepyris. ' Oriolus. 



Thamnophilus. Bill hooked at the tip. Myothera. 



(413.) Nothing'can be more perfect than the parallel 

 analogies resulting from comparing these two groups ; 

 and yet, as we have elsewhere demonstrated*, although 

 the divisions and analogies in both these columns, taken 

 separately, are correct, yet both are nevertheless disposed 

 falsely. Here then is a group which has undergone two 

 tests, in the first instance, it has a verisimilitude of being 

 truly circular, and then, being compared with an adjoin- 

 ing group, it is found to possess parallel analogies thereto, 

 and yet the great error of its composition remains to 

 be detected. How then are we to proceed in our pro- 

 cess of verification ? or how can a false circle be distin- 

 guished from a true one ? It is here that the third test 

 we have intimated, namely, the definite system of vari- 

 ation, must be resorted to, as a last and final criterion of 

 the true value of all groups, supposed to be natural. 



(414.) Now, the principles by which all the vari- 

 ations of form throughout the class of birds are regu- 

 lated may be thus concisely stated: First, we have, 

 in the typical form, a union of the greatest number of 



* Northern Zoology, vol. ii. p. 164. 



