evening song for an hour or longer at short intervals, and rendering it so perfectly that I was greatly surprised to hear it; for a thrush or other songster at the same period of life, when attempting to sing, only produces a chirping sound. The early singing of the oven-bird fledgling is important, owing to the fact that the group it belongs to comprises the least specialized forms in the family. They are strong-legged, square-tailed, terrestrial birds, generally able to perch, have probing beaks, and build the most perfect mud or stick nests, or burrow in the ground. In the numerous tree-creeping groups, which, seem as unrelated to the oven-bird as the woodpecker is to the hoopoe, we find a score of wonderfully different forms of beak; but many of them retain the probing character, and are actually used to probe in rotten wood on trees, and to explore the holes and deep crevices in the trunk. We have also seen that some of these tree-creepers revert to the ancestral habit (if I may so call it) of seeking their food by probing in the soil. In others, like Dendrornis, in which the beak has lost this character, and is used to dig in the wood or to strip off the bark, it has not been highly specialized, and, compared with the woodpecker's beak, is a very imperfect organ, considering the purpose for which it is used. Yet, on the principle that "similar functional requirements frequently lead to the development of similar structures in animals which are otherwise very distinct"--as we see in the tubular tongue in honey-eaters and humming birds--we might have expected to find in the Dendrocolaptidae a better imitation of the woodpecker in so variable an organ as the beak, if not in the tongue. Probably the oven-birds, and their nearest relations--generalized, hardy, builders of strong nests, and prolific--represent the parental form; and when birds of this type had spread over the entire continent they became in different districts frequenters of marshes, forests, thickets and savannas. With altered life-habits the numerous divergent forms originated; some, like Xiphorynchus, retaining a probing beak in a wonderfully modified form, attenuated in an extreme degree, and bent like a sickle; others diverging more in the direction of nuthatches and woodpeckers. This sketch of the Dendrocolaptidae, necessarily slight and imperfect, is based on a knowledge of the habits of about sixty species, belonging to twenty-eight genera: from personal observation I am acquainted with less than thirty species. It is astonishing to find how little has been written about these most interesting birds in South America. One tree-creeper only, Furnarius rufus, the oven-bird _par excellence,_ has been mentioned, on account of its wonderful architecture, in almost every general work of natural history published during the present century; yet the oven-bird does not surpass, or even equal in interest, many others in this family of nearly three hundred members.