Traces of what appear to be pre-spiracular clefts exist in the embryos of various forms. Perhaps the most remarkable of these is to be found in the larval Crossopterygian,[12] and apparently also in _Amia_[13] at least, amongst the other ganoids, where a pair of entodermal pouches become cut off from the main entoderm and, establishing an opening to the exterior, give rise to the lining of the cement organs of the larva. Posteriorily there is evidence that the extension backwards of the series of gill clefts was much greater in the primitive fishes. In the surviving sharks (_Chlamydoselachus_ and _Notidanus cinereus_), there still exist in the adult respectively six and seven branchial clefts, while in embryonic Selachians there are frequently to be seen pouch-like outgrowths of entoderm apparently representing rudimentary gill pouches but which never develop. Further evidence of the progressive reduction in the series of clefts is seen in the reduction of their functional activity at the two ends of the series. The spiracle, even where persisting in the adult, has lost its gill lamellae either entirely or excepting a few vestigial lamellae forming a "pseudobranch" on its anterior wall (Selachians, sturgeons). A similar reduction affects the lamellae on the anterior wall of the hyobranchial cleft (except in Selachians) and on the posterior wall of the last branchial cleft. A pseudobranch is frequently present in Teleostomes on the anterior wall of the hyobranchial cleft, i.e. on the inner or posterior face of the operculum. It is believed by some morphologists to belong really to the cleft in front.[14] _Phylogeny._--The phylogeny of the gill clefts or pouches is uncertain. The only organs of vertebrates comparable with them morphologically are the enterocoelic pouches of the entoderm which give rise to the mesoderm. It is possible that the respiratory significance of the wall of the gill cleft has been secondarily acquired. This is indicated by the fact that they appear in some cases to be lined by an ingrowth of ectoderm. This suggests that there may have been a spreading inwards of respiratory surface from the external gills. It is conceivable that before their walls became directly respiratory the gill clefts served for the pumping of fresh water over the external gills at the bases of which they lie. _Lung._--As in the higher vertebrates, there develops in all the main groups of gnathostomatous fishes, except the Selachians, an outgrowth of the pharyngeal wall intimately associated with gaseous interchange. In the Crossopterygians and Dipnoans this pharyngeal outgrowth agrees exactly in its mid-ventral origin and in its blood-supply with the lungs of the higher vertebrates, and there can be no question about its being morphologically the same structure as it is also in function.