quite impossible, where so many dialects probably exist, and where so little is known about any one of them, to fix any standard. I would, therefore, suggest that comparisons should be made under as diverse conditions as possible—that is to say, between very wet and very dry districts, or between districts inland and districts on the coast. If it can be proved that this variation exists among certain migratory species immediately on arrival in this country, it will be necessary to follow them into their winter quarters. For, supposing a dialect is inherent in any one given species (which at first seems almost incredible), we should expect to see some signs of it in their said winter quarters. On the other hand, if we follow them and again find new dialects and new gradations of tone, or if we find on their arrival in this country that there is no immediate variation, but that it increases as the season advances, we shall have strong evidence that in some measure at least it is directly due to climatic influence.
I have shown that so many and such distinct variations do exist, and it seems only reasonable to expect that some of these variations, amongst those species which are resident, will become hereditarily attached to the male sex—for if they did not it would be subversive of the theory of sexual selection, a theory which must be admitted by all those who have studied certain species, in whom the vocal powers are excessively developed while courting—consequently species with a certain variety of song will exist in a small body and often breed together, and as a result the development from a dialect to specific song must in time ensue. It may be argued that it is impossible for a variety of song amongst individuals of any one species to have any connection with the origin of song in separate species; but I can see no more difficulty in believing, except to those, if there are any, who still look upon species as immutable, that through the vast ages that have lapsed, during which species have developed, a specific song may have become attached to a certain species through the action of sexual selection on varieties resulting from climatic influence, than I can in believing that species themselves have been evolved.
When we reflect on these variations of song, we can easily understand what mistakes have arisen, and probably will arise,
