Abstract.-The porpoise, an air-breathing mammal whose habits are entirely aquatic, presents special problems of respiration, sleep, and anesthesia. These problems have been studied in three species, Tursiops truncatus, Lagenorhynchus obliquidens, and Phocoenoides dalli.The respiratory rhythm is unusual in that there is an extended pause-an apneustic plateau-between periods of expiration and inspiration. This rhythm has been observed under waking conditions, during sleep, and also when the animal is anesthetized. Two kinds of sleep behavior have been identified in Tursiops and Lagenorhynchus. In one of these, which is a light phase, the animal rests on the tank bottom for short periods, up to perhaps four minutes, and then comes to the surface to breathe. In a deeper phase the animal maintains itself at the surface for extended periods, breathing in an automatic fashion. In Phocoenoides no sleep behavior has been observed at any time.

In the dolphin Tursiops truncatus, the basilar membrane over its course from basal to apical ends shows a systematic variation in width and in the manner and rigidity of its suspension.The suspension is by bony laminae on both the outer and inner edges in the basal region, and by progressively less rigid attachments in the more apical regions, until near the apex the membrane is held only by connective tissue.The basilar membrane shows an unusual variation in width of 14 times, ranging from an astonishingly small value of 25 pm at the basal end to a value toward the apex of about 350 pm. This structural variation is in harmony with the extension of sensitivity of the dolphin ear into the very high frequencies, and suggests unusual capabilities of pitch discrimination in this high range.In a preceding article (1) we outlined the present status of our knowledge concerning the detailed anatomy of the dolphin cochlea, and reported some recent observations on the general morphology of the inner-ear structures of Tursiops truncatus. Here we present results on the form and suspension of the basilar membrane. METHODAs described in our earlier article, the specimens were perfused intravitally with a fixative that in the more successful experiments produced an excellent preservation of cochlear elements. In all specimens, the basilar membrane was well maintained. Seven of these specimens were studied in detail for this report.The first step in the examination of these ears was a graphic reconstruction of the cochlear spiral by Guild's method (2), which provides a representation to scale of the structures as projected on a plane perpendicular to the cochlear axis. Such a spiral diagram was shown in Fig. 2 of the preceding report (1).

A study of the cochlear hair cells in Tursiops truncatus showed 3451 inner and 13,933 outer hair cells, for a total of 17,384. This total is of the same order of magnitude as the value of 14,975 for the human ear. Determination of the ganglion cell population for the dolphin gave a total of 95,004 cells, which is about three times as many as in man.The large number of hair cells in the dolphin ear suggests a high order of auditory proficiency in general, and especially a marked ability of tonal differentiation. The large ratio of ganglion cells to hair cells suggests unusual capabilities in the utilization of auditory information.In two previous reports (1, 2), we described the general morphology of the dolphin cochlea and gave particular consideration to the form of the basilar membrane and its manner of suspension. This paper deals with the numbers of hair cells and ganglion cells and their distribution along the cochlea. These features are of interest because they relate to questions of the specificity with which the action of sounds on the ear can be represented at the level of hair-cell stimulation and in the initial involvement of the auditory nervous system. THE HAIR CELLSEarly in the study of the anatomy of the vertebrate ear, it was observed that as we go from simpler to more advanced forms, we find a large increase in the size of the auditory papilla and in the numbers of its hair cells. Among the mammals, Retzius in 1884 found this increase in numbers in the series from the rabbit through the cat to man. It was readily inferred that this progression is related to the characteristics of these different ears in tonal range and frequency discrimination.In determining the size of the hair-cell population, Retzius (3) used a method that had been developed earlier by Waldeyer (4) before the development of serial sectioning. It requires much patience and skill, but provides an excellent orientation and appreciation of the forms of anatomical elements.Waldeyer in 1872 examined the human cochlea by this procedure; his work was followed by that of Krause in 1876. These investigators exposed the organ of Corti in one region of the cochlea and counted the outer pillar cells over some convenient distance, from which they obtained the average spacing of these elements. They assumed that the hair cells (that were much more difficult to see under their conditions) had the same spacing as the outer pillar cells. Then, they measured the length of the auditory papilla (or the basilar membrane); by dividing this length by the spacing, they obtained an estimate of the number of elements in a row along the cochlea. Finally, by multiplying this figure by the number of rows, they found the total number of hair cells.Retzius repeated these measurements with some refinements of technique and with the observation that in man there are only three rows of outer hair cells over most of the cochlea, whereas his predecessors had thought that there were four rows throughout. Retzius found four rows only in the apical region...

Abstracts.-Measurements were made of the cochlear potentials in three specimens of the giant sea turtle, Chelonia mydas, in response to both aerial and mechanical stimulation of the ear. The results show maximum sensitivity in the region of 300 to 400 Hz, with a rather rapid decline in sensitivity for lower and especially for higher tones. The upper limit for the observation of cochlear potentials without injury is 2000 Hz, and a practical limit of usefulness of this ear is probably about 1000 Hz.The presence of hearing in turtles is now well established through both behavioral and electrophysiological observations. However, these observations have been limited to the small and readily available species, and nothing definite is known about the giant marine forms. An early attempt by Foa and Peroni to record auditory potentials in large specimens of Thalassochelys caretta gave no useful results.1 1\Ieasurements of auditory sensitivity in terms of the cochlear potentials were carried out by Wever and Vernon in 1956 on four of the small common species: Chrysemys p. picta, Clemmys insculpta, Pseudemys scripta, and Terrapene c. carolina, all belonging to the family Emydidae.2-4 More recently, the hearing of one of these species, Pseudemys scripta, was tested by a behavioral method by Patterson.5The patterns of auditory sensitivity obtained for all these animals, and by both methods on P. scripta, were closely similar.6 In every instance a fairly high level of sensitivity was observed for low tones, with the best responsiveness in the region of 200 to 700 Hz. For tones below 100 Hz the sensitivity was found to decline slowly in the two species in which tests were made in this region, and for tones above 700 Hz the sensitivity of all species was found to fall off with great rapidity, so that beyond 3000 Hz no tests could be made without injury to the ear.The present experiments were carried out on the giant sea turtle, Chelonia mydas. The specimens were obtained from commercial sources in Hawaii, and thus no doubt represent the Pacific subspecies. Measurements of sensitivity to sounds were made by the cochlear potential method in 3 animals, whose weights were 50, 68, and 75 pounds. Several additional animals were used for anatomical and surgical studies. All the potential measurements were made on the left ear.Procedure.-The animals were held in a saltwater pool until needed for the experiments. While in the pool, they were observed to remain on the bottom at a depth of 8 ft most of the time, and they came to the surface periodically to breathe. The smaller specimens usually surfaced at intervals of ten minutes 884