100 ME. W. a. EIDEWOOD OK THE HTOBEANCHIAL 



branchial basket-work fits underneath the auditory capsule so' 

 closely that considerable care is required to separate the two ; 

 but in this third stage the whole of the branchial skeleton lies in 

 the floor of the pharynx. 



The most interesting feature o£ this stage is the formation of 

 the thyrohyals, which may be seen arising from the hind edge of 

 the median basal plate of cartilage, as a paired outgrowth per- 

 fectly independent of the four branchial arches (PI. 11. fig. 3, f). 

 This is universally accepted as the mode of origin of the thyro- 

 hyals in the tongued Anura. It was clearly set forth in the 

 works of Cuvier (8), St.-Ange (36), and Duges (10), was con- 

 firmed by Parker (32) in his exhaustive treatise on the develop- 

 ment of the skull of the common frog, and has recently been 

 corroborated by Schulze (38) and Graupp (14). These thyrohyals 

 as yet show no trace of ossification, and are quite free from the 

 laryngeal skeleton. Here, then, is conclusive evidence of the 

 secondary nature of that connexion between the hyobranchial 

 and the laryngeal skeleton, which distinguishes the Aglossa from 

 all other Amphibia. The fourth branchial arch has shrunk 

 considerably since Stage II., especially towards its median end, 

 so that not only is it now no thicker than the third arch ; but it 

 no longer underlies the larynx. In fact, although the larynx has 

 not shifted in position relatively to the branchial clefts, there has 

 appeared a space between the fourth branchial arch and the 

 larynx. Into this space the thyrohyals grow. 



Owing to paucity of material I have been unable to determine 

 the mode of origin of the great alge of the adult hyobranchial 

 skeleton ; but it is highly probable that, as in Pipa, these are 

 the result of secondary outgrowths from the undivided hypo- 

 branchial part of the branchial skeleton (PI. 11. fig. 3, Jih), and 

 not, as might be suggested by a comparison of fig. 3, PL 11. and 

 fig. 1, PI. 8, of the persistence of such part of the branchial 

 skeleton as still remains, under a condition of closure of the 

 clefts. The ceratohyals, still massive, are fated to undergo a 

 certain amount of absorption. The absorption, however, is not 

 excessive, and the lamellar expansions of cartilage on the 

 internal and external edges of the anterior cornua of the adult 

 clearly owe their presence to the fact that the absorption of 

 cartilage is less complete than in most Anura. The laryngeal 

 skeleton has altered but little since Stage II. The arytenoid 

 cartilages (PL 11. fig. 10, ar.) are very large in proportion. They 



