582 DR. W. G. RIDEWOOD ON THE HTOBRAlfCHIAL [l^faj IS, 



carefully freed from the skull. The cesopliagus was then cut in 

 front of the liver, and the Mhole of the floor of the pharynx thus 

 liberated. The heart and larynx were then dissected awny, the 

 gills, muscles, and mucous membrane removed, and only sullicient 

 connective tissue left to keep the skeletal parts in their natural 

 relations. No staining reagents were employed, and the whole of 

 the dissection was performed binder a simple microscope of a 

 magnification of ten diameters, in powerful reflected light. The 

 series of larvae examined vvas very extensive, hut it will suffice to 

 select eight stages only of these for description. The important 

 changes in the hyobrancliial skeleton do not begin until the tail is 

 reduced to about one-third of its maximum length, but they then 

 proceed with considerable rapidity. In fact, two larvse exactly 

 similar in external appearance and having the merest stump of the 

 tail remaining may show considerable differences in the structure 

 of the hyobranchial skeleton. In such cases it is necessary to make 

 the dissections first and to arrange the specimens in series aftei-- 

 wards. The youngest tadpole I have been able to examine is one 

 with the hind limbs just appearing as buds on the surface of the 

 body, but from this onwards, as far as the adult condition, the 

 series is as complete as could be wished. It should be noted that 

 the hyobranchial skeleton of the first seven stages is drawn from 

 the dorsal side, in order to show the spicula and the articular ends 

 of the hyoidean cornua. Tigs. 8 and 9, however, exhibit the 

 ventral surface, because the spicula have by this time disappeared, 

 the hyoidean cornu has no longer an articular surface, hut is 

 firmly adherent to the auditory capsule, and because the interesting 

 ventral ossification is only to be seen in this view. In order to 

 facilitate comparison the figures are not drawn to the same scale, 

 but as nearly as possible of the same absolute size. The approxi- 

 mate magnification is given in each case. The introduction of the 

 mandible into the figures serves to show not only the gradually 

 increasing size of the gape, but also illustrates the interesting 

 manner in which the distal extremities of the hyoidean cornua get 

 pushed farther and farther back as the mandibular rami elongate 

 posteriorly. Throughout this contribution, although ostensibly 

 devoted to a consideration of the hyobranchial skeleton, I have 

 purposely avoided all reference to the columella auris and 

 stapedial cartilages. 



Stage 1. Distance from snout to root of tail, 13 mm. Length of 

 tail, 16 mm. Length of hind limhs, 1 mm. (Plate XXXV. fig. 1.) 



There is nothing very remarkable about the hyobranchial 

 skeleton of Felodytes in its early stages, since it conforms tolerably 

 well with what might be considered the normal for tadpoles in 

 general. There is but a single unpaired median constituent, and 

 not two, as, for instance, in Ahjtes. Although Parker (12) speaks 

 of this skeletal part of the Anuran larva as the " basibranchial " 

 and Stanuius (19. p. 64), Hoffmann (7. p. 45), and Gaupp (5) 

 apply to it the non-committal name "copula," the general 



