IN THE PARAGUAYAN LEPIDOSIKEN, ETC. 371 



In the present communication it is not intended to refer to the obscure problem of 

 the phylogenetic origin of the Dipnoi themselves. An exhaustive and suggestive 

 discussion of this question is to be found in the previously-cited paper by Dollo. There 

 is, however, one point which may be referred to as having a direct bearing on this 

 problem. 



In common with several other morphologists, Dollo regards the Crossopterygii as the 

 ancestral stock from which the Dipnoi have been derived, and bases his conclusion on 

 the prevalence in certain fossil Dipnoi of characters common to the two groups, such 

 as, for example, the existence of lobate paired fins, the presence of jugal plates, the 

 ventral position of the narial apertures (in some Crossopterygii), and the ganoid 

 investment of certain of the cranial bones. The obvious difficulty presented by 

 the hyostylisra of the Crossopterygii, and the pronounced autostylism of the Dipnoi, is 

 met by the bold suggestion that the latter is an adaptive modification associated with 

 the development of the peculiar Dipnoid type of dentition, which has also conditioned 

 the suppression of the usual dentigerous bones, the premaxillse, the maxillse, and the 

 dentaries. In support of this suggestion, Dollo emphasizes the association of auto- 

 stylism with massive palatal teeth in the Holocephala and the incipient autostylism of 

 Cestracion with the cochliodont dentition of that Selachian. 



With regard to Dollo's theory of the adaptive character of autostylism, and its 

 independent acquisition by widely different groups of Fishes, it may be remarked that 

 the autostylism of the Cyclostomes and Amphibia still remains as a difficulty in the 

 way of its acceptance, inasmuch as in neither group can the prevalence of autostylism 

 be explained as due to the peculiarities of an exceptionally-developed dentition. It is 

 nevertheless quite possible, and in my opinion even probable, that other factors besides 

 peculiarities of dentition may condition the evolution of autostylism, and one of these 

 may be the development of a suctorial mouth supported by a more or less elaborate 

 system of labial cartilages. In the latter case the advantage derivable from the fixation 

 by autostylism of the primary skeletal elements of the upper jaw may conceivably 

 be quite as great as that gained in the case of a massive dentition. But, apart from 

 the reasons advanced by Dollo, there are other considerations which suggest the 

 derivation of the autostylism of the Dipnoi from a hyostylic type of skuU. 



As shown by Huxley [15], the upper extremity of the hyoid arch (cerato-hyal) in 

 Ceratodus is connected with the suspensorial cartilage through the intervention of a 

 cartilaginous element which he regards as the representative of the hyomandibular 

 bone or cartilage of other Fishes. A similar interpretation may perhaps be assigned to 

 the upper division of the hyoid arch in Chimcera, which, like the hyomandibular of 

 Elasmobranchs, carries the dorsal series of hyoidean rays. Assuming the correctness 

 of this interpretation of the nature of these cartilages, and that they are really homo- 

 dynamous with the hyomandibular element of the majority of Fishes, au obvious 

 inference is at once suggested. In practically all existing Elasmobranchs and 

 Teleostomi, or at aU events with only a single exception {Notidanus), the hyo- 



