tion, and are exposed to essentially like environmental conditions." Is not the manner of song modification as much "identical material" of the Brown Towhee as is his brown coat or his mincing step as he forages in the grass ? Is it not as natural that two individuals of a species should behave alike in improving as in possessing a song ? The facts impressing me as significant because of their appearance in both of our towhee songs, are as follows:
- The fact of elaboration.
- The fact of elaboration through imitation.
- The fact of elaboration through adding the imitated syllables at the end.
- The fact of the comparative musical quality of the imitated syllables (a
contrasted with the hard metallic quality of the notes in the "bouncing" song).
Certain ideas, of not too speculative a character, I hope, to be of some value, occur to me in connection with each of these four points.
1. The fact of elaboration.—It would seem that all birds who possess songs possess a tendency to elaborate or improve their, songs; that is, to render more perfect their songs in the number, kind, and arrangement of notes. The finest singers among birds (so considered from the human point of view, and so too in actuality, I believe) are the producers of comparatively prolonged and complex musical effects. The evolution of bird song evidently proceeds from simplicity to complexity.
The simplest songs are those sung in short set form—running automatically in pro-established grooves. Within this definition, the simplest are those containing only one kind of sound. Finally, these two essentials being complied with, the simplest are those in which the one kind of sound involved is derived from the call- or alarm-note of the singer. Charles A. Witchell describes (1896, pp. 49-50) the songs of some fifteen birds, and remarks (p. 53), "In all the above-mentioned British species, and in some of their allies, which represent many avian races, the males court the females partly by the repetition of notes which we observe to be employed in other circumstances as call-notes; and in some species these notes are repeated so rapidly that a phrase is constructed. But some species have never advanced beyond the mere repetition of their call-notes." He expresses the belief (p. 58) that "songs were, at first, mere repetitions of call-notes, or possibly of defiance-cries, which have since been more rapidly uttered and varied, with the result that novel strains have been slowly developed." Among American birds two species occur to me off-hand as possessing songs more or less of this type. One is the Linnet, whose zig-zag, twittery warble is found on analysis to be composed partly of the call-note, or chirp, of the species. The other is the object of our present study, the Brown Towhee, whose normal song is composed wholly of repetitions of its alarm-note.
The typical "bouncing" song of the Brown Towhee, though according to our definition very simple, is doubtless even at that an elaboration of something still more simple, which in turn we must trace back through imperceptible stages to its humble origin, the alarm-note. The alarm-note (in the Brown Towhee as in other song-birds) must have existed before the song. Without bothering ourselves here about the origin of alarm- and call-notes (whether to be found in spasmodic muscular contractions acting. upon the trachea in moments of excitement or of effort, or to some necessity of emotional. expression for purposes of social control [see Wallace Craig, 1908]) we can feel pretty sure that the song is the most advanced form of expression in the bird's vocal his-
