584 DR. W. G. RIDEWOOD OX THE HYOBRANCllTAL [May 18, 



cartilage. On the external edge of the first ceratobrancliial these 

 processes are of considerable length and form a series of five or 

 six. tSimilar, but more irregular, cartilaginous processes stand out 

 from the postero-internal border of the fourth ceratobranchial and 

 from the external boi'der of the commissural cartilage. If, now, 

 these ii'regular finger-shaped pi'ocesses of the first ceratobranchial 

 were conflueut laterally so as to form a continuous curved wall, 

 and if the same coalescence occurred wiih regard to the processes 

 of the commissural cartilage and fourth ceratobranchial, the whole 

 branchial skeleton would assume the form of two approximately 

 hemispherical cartilaginous basins perforated at the bottom by 

 three oblique slits — a form of branchial skeleton which is peculiar 

 to Xeno2ms (16). Which of the two forms is the more primitive, 

 that with the irregular palisade, or that with the continuous 

 marginal wall, it is dithcult to say ; but I am inclined to regard the 

 branchial skeleton of the early Xenopns larva as more primitive 

 than that of any other Anuran larva yet described '. The fourth 

 branchial cleft (the last of the three perforations in the branchial 

 cartilage) is shorter than the two preceding. The distal end of 

 the third ceratobranchial is exceptionally broad, and is character- 

 ized by the constant presence of either a deep notch or a foramen. 

 The proximal end of the first ceratobranchial is bi'oad, and 

 merges indistinguishably into the cartilage of the bypobranchial 

 plate. It bears a backwardly directed spicule of cartilage 

 (sp\ fig. 1). The second ceratobranchial (c6r-, fig. 1) is boimd 

 by connective tissue to the posterior surface of the first cerato- 

 branchial, to the median side of this spicule. The proximal part 

 then curves upwards and backwards, and giving off a backwardly 

 directed spicule of its own, diminishes in thickness and becomes 

 continuous with the cartilage of the proximal end of the third 

 ceratobranchial, just where the latter unites with the bypobran- 

 chial plate ■-. The third ceratobranchial curves over in a similar 

 manner, and has its own spicule {sjf, fig. 1). The fourth 

 ceratobranchial is broader and far more irregular in shape than 

 the other three. It curves up, is confluent with the bypobranchial 

 plate, and sends backwards a curved cartilaginous process, which 

 may possibly represent the spicule of this arch. It should be 

 noted that the whole of the cartilage of the branchial plate is 

 continuous throughout. This continuity is obvious at the distal 

 ends of the arches, but, on account of the sharp curvature of the 



^ Schulze (18. p. 11) and Naue (9. p. 14) seem to regard the front wall 

 belonging to the first ceratobranchial as a thin, continuous, concave plate of 

 cartilage in the forms of Anura which they respectively studied. The plates 

 are probably subject to a great amount of individual variation, but in none of 

 the numerous larva^ which I have examined {Rana, Biifo, Ali/tcs, Pelobates, 

 Pelodytes, Hyld) have I found the wall nearly so continuous and entire as in 

 Xenopus. 



^ I fail to see the force of Gaupp's argument (5. p. 403) for considering the 

 parts of the first two branchial bars lying internal to their spicules as belonging 

 to the bypobranchial plate rather than to the ceratobranchials (branchialia) 

 themselves. 



