very highly conventionalized derivatives of the alligator radical, the meandered line representing the body of the creature and the scalloped hooks the extremities (Fig. 238). The two bands upon the base consist of geometric figures, the origin of which cannot be definitely determined, although they also probably refer to the alligator.

Fig. 238. Conventional design on drum shown in Fig. 237, composed of alligator derivatives.

In the collection there is a minute toy drum of the same general shape, and the same form reappears in some of the whistles, in one of which ([Fig. 247]) the skin head and its fastenings are all carefully reproduced in miniature. The immediate original of this particular form of drum was probably made of wood. A drum, recently brought from Costa Rica was made by hollowing out a cylindrical piece of wood and stretching a piece of snakeskin across the top. The shape is nearly identical with that of these earthen specimens.

[ Wind instruments.—]

Earthenware wind instruments are found in considerable numbers and are associated with other relics in the tombs. Nearly all are very simple in construction and are limited in musical power, receiving and perhaps generally deserving no better name than whistles or toys. A few pieces are more pretentious and yield a number of notes, and if operated by skilled performers or properly concerted are capable of producing pleasing melodies. It is not difficult to determine the powers of individual instruments, but we cannot say to what extent these powers were understood by the original owners, nor can we say whether or not they were intended to be played in unison in such a way as to give a certain desired succession of intervals. There are, however, in a large number of these instruments a uniformity in construction and a certain close correspondence in the number and degree of the sounds that indicate the existence of well established standards. It does not appear absolutely certain to me that the system of intervals was made to conform to that of any known scale; but a difficulty arises in attempting to determine this point, as most of the pieces are more or less mutilated. We find also that the note producible by any given stop is not fixed in pitch, but varies, with the force of the breath, two or even three full intervals. As a result of this a glide is possible to the skilled performer from note to note and any desired pitch can be taken.

In material, finish, and decoration these objects do not differ from the ordinary pottery. A majority belong to the alligator group. The size is generally small, the largest specimen being about eight inches in length. The shapes are wonderfully varied and indicate a lively imagination on the part of the potter. Animal forms prevail very decidedly, that of the bird being a great favorite. In many cases the animals copied can be identified, but in others they cannot—perhaps from our lack of knowledge of the fauna of the province, perhaps from carelessness on the part of the artist or from the tendency to model grotesque and complicated shapes. The following creatures can be recognized: men, pumas, ocelots, armadillos, eagles, owls, ducks, parrots,

several varieties of small birds, alligators, crabs, and scorpions. Vegetal forms, excepting where in use as instruments or utensils, as reeds and gourds, were not copied. In the National Museum collection there are two tubular pipes, probably modeled after reeds, and another resembles a gourd in shape. The construction of the whistling apparatus is identical in all cases and corresponds to that of our flageolets (see sections, Figs. 240 and 242). Plain tubes were doubtless also used as whistles, and all utensils of small size, such as needlecases and toy vases, can be made to give forth a note more or less shrill, according to the size of the chamber. The simplest form of whistle produces two shrill notes identical in pitch. The shape is double, suggesting a primitive condition of the tibiæ pares of the Romans. The parts are pear or gourd shaped, are joined above and below, and have an opening between the necks. The two mouthpieces are so close together that both are necessarily blown at once. The note produced is pitched very high and is extremely penetrating, not to say ear splitting, making an excellent call for the jungles and forests of the tropics. A small specimen is presented full size in Fig. 239, and the section in Fig. 240 shows the relative positions of the mouthpieces, air passages, vent holes, and chambers.
Fig. 239. Double whistle, lost color ware—1/1. Fig. 240. Section of double whistle.