662 BRITISH BIRDS. 



its allies, objects very strongly to my having degraded these races of Wrens 

 to the rank of subspecies. Since the article was written I have had an 

 opportunity of examining many more specimens, and am still more con- 

 vinced that in a large series from any one locality the extremes would be 

 found to be indistinguishable from extreme examples belonging to the 

 nearest allied subspecies. The main characters relied upon to distinguish 

 the various races of Wren are the general colour of the plumage (which varies 

 from brownish grey to rich russet-brown), the presence or absence of dark 

 bars across the feathers of the back and breast, and the size of the bill and 

 feet. The following table will point out the variations of the subspecies 

 in these respects : 



Pale colour 



It will be at once seen that the slight differences between these races bear 

 little or no relation to geographical distribution. It is rather a remarkable 

 fact that the nearest ally of the St.-Kilda Wren should be found on a group 

 of islands subject to the same meteorological conditions, and in consequence 

 equally characterized by the absence of trees. Similarly the Wren of 

 Algeria scarcely differs from that of Turkestan. There can be little doubt 

 that the races of Wren, like those of the Creeper, the Nuthatch, the Marsh- 

 Tit, and the Coal Tit, appear to differ in colour according to climate and 

 not according to geographical distribution, except so far as it happens to 

 be connected with climate. 



The Wren is one of those widely distributed and variable species of 

 birds which are the despair of nomenclators. No fewer than a dozen 

 races are known, and doubtless a dozen more remain to be discovered. A 

 bird which is only migratory in the extreme north of its range and is found 



