THE AUTUMN SONG OF BIRDS.
The songs of birds are worth investigating; but before progress can be made in the knowledge of the why and wherefore of these songs, we must ascertain how and when they occur. The last particular is especially important. It is very well to attribute the songs of birds to an erotic origin; but that will hardly account for the Robin and Starling recommencing in July or the first few days of August. Nor will it account for autumnal songs which are preceded by a period of silence (e.g. the Chiffchaff), or which are followed by a silence, which is not the case with the Robin and Starling. The September songs of Willow Wren and Chiffchaff are so exceedingly few and far between, as compared with the spring songs, that they may very probably proceed from birds that did not breed in spring, or whose nests were destroyed.
My particular reason for calling attention to this theme is that my own observations seem to conflict with some other records. This may be due to the fact that I have always been "an early bird"; while other observers with less exacting avocations may be more of midday or evening observers. When articled and subsequently in a practice at Stroud, where most of my observations were made, I never loafed after birds during office hours, but was out on nearly half the fine mornings from 6 or 6.30 till 9 o'clock a.m.
In a paper on the autumn song of birds (Zool. 1894, p. 410) Mr. O.V. Aplin says that the Willow Wren (after being silent from mid-June) strikes up again about the second week in August. The words "strike up" are, however, also applied to the Robin and Starling in November or October. Mr. Aplin has assured me that the remark does not, in the case of the latter birds, mean commencing to sing, but the employment of a new style of song. As, however, in the same paper he has a special reference to the Starling as singing in October (why October,
