X INTRODUCTION, 



tho probcabilities are tbat it ia a specific difference. On the other 

 hand, if the two forms are so closely allied that they interbreed 

 and produce fertile offspring, which again interbreed, we may draw 

 two inferences — first, that the difference between the two forms 

 is only subspecific, and, second, that between the two extremes 

 must be an infinite gradation of intermediate forms. If the two 

 forms have different areas of geographical distribution (which is 

 usually the case), the intermediate ones will be found where the 

 two areas overlap. If, from any cause, the species should be ex- 

 terminated in the overlajiping areas, and the causes which produced 

 the variations of the two forms still continue, the peculiarities 

 of each become emphasized, until they become so far separated, 

 that should their areas of distribution again overlap they wiU 

 nevertheless not interbreed, and the two species may bo considered 

 to be completely segregated. Under these circumstances, I have 

 preferred to retain the binomial nomenclature for each of the ex- 

 treme forms, reserving the trinomial name for the intermediate 

 ones, uniting the two specific names by a hyphen, and placing the 

 name of the bird first which it most resembles. Sterile hybrids 

 maj- be denoted also by a trinomial name, the two specific names 

 being united by a cross, the name of the male taking the pre- 

 cedence. 



N'S hilst I have endeavoured to bring the treatment of genera and 

 of specific and subspecific forms into harmony with the latest 

 discoveries of science, and therefore may bo considered an orni- 

 thological revolutionist by those who have not yet accepted the 

 modern theories of evolution, upon questions of nomenclature I have 

 adopted conservative principles. The former involve questions of 

 fact, in which it is necessary to march with the progress of research ; 

 the latter are merely questions of names, in which it must bo obvious 

 to the merest tyro that change is in itself an evil. It appears 

 to me that the recent attempts to carry out certain parts of the 

 Stricklandian code to an extreme that may not inaptly be called 

 revolutionary, has been productive of much confusion and un- 

 necessary novelty. This I have endeavoured studiously to avoid by 

 retaining the names in familiar use for many common birds, even 

 where those names are not the oldest. I have only rejected them 

 in cases where the name has been so frequently misapplied to 



