98 MB. W. Qo EIDEWOOD OK THE HTOBEANCHIAL 



small cartilage between the ceratohyals is undoubtedly the basi- 

 byal, and that the larger, and more posterior, unpaired cartilage, 

 which is of more constant occurrence, has rather the value of a 

 basibranchial. Schulze (38) still holds to the original view of 

 Duges (10), that this latter is the basihyal (basihyoid). 



The larynx at this stage is quite diminutive in size, and lies 

 above the level of the internal reflected margins of the branchial 

 cartilages. The width of the larynx is hardly more than that of 

 one of the bronchi, and the length of the glottis is about half 

 the width of the larynx. The bronchus is not sharply marked off 

 from the lung, but appears rather as a tubular non-sacculated con- 

 tinuation of it. The bronchus and lung are about equal in length. 

 As might be expected at this early stage, the lungs are very small, 

 and the total length of bronchus and lung together is not more 

 than the maximum transverse diameter of the branchial skeleton. 

 Arising from the dorsal surface of the anterior end of each 

 bronchus is a curious thin-walled sac (Plate 11. fig. 1, p) which 

 runs up laterally to the oesophagus, and is lodged beneath the 

 ribs (see p. 120) of the anterior vertebrae. The significance of 

 these sacs it is difficult to estimate : they are probably to be 

 regarded as accessory lobes of the lung, since their walls so closely 

 TBsemble the lung-tissue. They are present in all the larvae of 

 Xenopus that I have examined, but are altogether absent in the 

 adult. I have found nothing to correspond with them in either 

 the young or adult of Pipa. 



Stage II. (Plate 11, fig. 2.) Tadpoles with fore Imibs mode- 

 rately well developed, pigmented, hut not markedly angulate. 



By comparison with Stage I., the most important difference to 

 be noted in Stage II. is the reduction in size of the branchial 

 skeleton. The hyoid arch and mandible show no great change. 

 The curvature of the outline in the dorsal view of the branchial 

 skeleton is no longer seen, but each half is roughly five-sided. 

 The inflation is less conspicuous, and the basket is considerably 

 shallower. The overhanging fold at the anterior end is missing, 

 although the ledge projecting outwards from the applied mesial 

 surfaces still remains. The branchial clefts have not altered ; 

 and they serve to show that the absorption has been greatest 

 anteriorly. The first branchial arch (PI. 11. fig. 2, cl. 1), which 

 before fitted so closely into the concavity at the back of the 

 <5eratohyal, now barely overlaps the ventral ledge of that cartilage 



