162 Elliot on the Genus Dendrornis, [April 



examination of my material by any unbiased naturalist, would 

 permit no other conclusions to be arrived at than those expressed 

 in these pages. It is not pleasant to disagree with those working 

 in the same field, and wherever a shadow of a doubt existed, I 

 have preferred to give the benefit of that 'shadow,' if any could be 

 derived from it, to a fellow ornithologist, but when my material 

 caused me to believe that an error had been committed, I have 

 not hesitated to say so, with my reasons for such belief. 



I have found that measurements, taken as a basis of scientific 

 value in this group, are of little worth, as examples of the same 

 species vary in their dimensions even from the same locality, and 

 when such differences were considerable I have given the extreme 

 measurements of the specimens before me, beside those of the 

 example described. The length of the bill given is always that 

 of the exposed culmen. 



Knowing bv experience how extremely difficult it is to distin- 

 guish closely allied species of this genus by descriptions only, 

 and not hoping to be more successful in this respect than any 

 of those who have written upon this group, I have constructed 

 a key, containing distinctive characters, or the salient character 

 of each species, by the aid of which, I trust, little difficulty 

 will be met with by any one in deciding what particular 

 species he may have before him. Of course, to those who 

 consider that this genus should contain races, sub-races, etc., 

 not recognized in this paper, and not apparent to the author, 

 the key will only be useful up to a certain point, beyond which 

 they must provide their own conveyance for farther travel. I 

 have not attempted to quote what might be considered a full 

 synonymy for each species, as in many cases where the name 

 only was given, no description having been added, it was impos- 

 sible to determine with any degree of accuracy what was the 

 species intended. That errors were made, was shown in certain 

 instances, where an author in a subsequent paper had altered the 

 name previously given to his specimen, to that representing quite 

 a different species. It therefore seemed best, when no description 

 was given, and access to the specimen mentioned was not avail- 

 able, that in the great majority of cases no notice of its occurrence 

 should be taken, and that I should confine myself to such refer- 

 ences, regarding which there was no reasonable doubt as to what 

 species was mentioned. To cite an instance of the difficulty of 



