106 ME. W. G. EIDEWOOD ON THE HTOBRANCHIAL 



extent by the shortening of the branchial clefts before they 

 become broken open peripherally ; but it is more conclusively 

 shown by the fact that, while in Stages I. and II. the aortic vessels 

 made up of the four efferent branchial arteries riin alongside the 

 commissural or epibranchial cartilages that connect the outer 

 ends of the branchial arches, there exists in Stage III. a con- 

 siderable interval between the paired aorta and such parts of the 

 commissural cartilages as yet remain. This is not to be explained 

 by a widening or separation of the aortse, for the actual length 

 of the branchial vessels has not increased. The commissure 

 between the first two arches was already ruptured in Stage II. : 

 it has now completely disappeared. The second commissural 

 cartilage has severed its connexion with the second branchial 

 arch, but remains attached to the third. The third commissure 

 is still intact. The first branchial arch is undergoing reduction 

 at its proximal as well as at its distal extremity. The cartilage 

 at the junction of the first ceratobranchial and the hypo- 

 branchial plate (PI. 11. fig. 6, t) is very thin, and showed in two 

 of the specimens examined a distinct indentation in the edge, — 

 a foreshadowing of the dismemberment of the arch which is 

 completed in Stage IV. A considerable reduction is to be noticed 

 in thickness of the second branchial arch. The fourth arch, on the 

 contrary, is thicker than before, especially towards its proximal 

 extremity, where it is in continuity with the hypobranchial plate. 

 The latter plate has become perforated by a paired foramen (t') 

 lying immediately over the thyroid bodies. The cartilage situated 

 to the outer side of the foramina shows unmistakable signs of 

 absorption, but the front and hind portions of the hypobranchial 

 plate are increasing in thickness. 



The thickening at the posterior edge of the plate is highly im- 

 portant, because, as will be seen by comparison with the later 

 stages, it indicates the formation of the thyrohyals. The thyro- 

 hyals in jPtj)a do not grow out as distinct processes as they do in 

 the tongued Anura, and even in Xenopus, but are formed by an 

 addition of cartilage to the hind edge of the hypobranchial plate<r 

 The added cartilage also extends outwards behind the proximal 

 end of the fourth branchial arch, and this it is that causes the 

 apparent increase in thickness of the last arch. There is no 

 augmentation of the cartilage of the arch, but an addition to it& 

 posterior edge. This is indeed a most aberrant mode of forma- 

 tion of the thyrohyals, unparalleled in the Anura; but the 



