172 



TRANSACTIONS OF THE AMERICAN PHILOSOPHICAL SOCIETY 



turned "proboscis." This liypothesis would account for the following otherwise inexpli- 

 cable facts: (1) the nasal bones and mesethmoid of Petrocephalus are turned very sharply 

 downward as they would be if derived from those of more typical mormyrids; (2) the form 

 and detailed relations of the premaxills, maxillae and dentary strongly suggest the condi- 

 tions in typical mormyrids and the same is true of the form and relations of the opercular, 

 preopercular and hyopalatine series; (3) the hypothesis of a trunk curved convexly in front 



./'^••.. 



dssp '^^P^ boc /<^a 



P 



Petrocephalus bane 



brsfg 



Fig. 64. Pitrocephatus bane. After Ridewood. 



of the eyes and then quickly shortened would also seem to account for the very close 

 morphological relations of Petrocephalus to typical mormyrids in the characters of the 

 cranium and in the general body form; (4) the skull of Marcusenius psittacus is intermediate 

 between the forms with down-curved snouts and Petrocephalus. Its tiny jaws are partly 

 decurved but are much shortened antero-posteriorly, so as to lie beneath the obliquely- 

 placed nasals. 



Here then is an excellent example of the "irreversibility of evolution" in the sense 

 of Dollo's law, for the deeper morphological results of a former lengthening of the mid 

 portion of the skull and of the development of a trunk are still evident even after the 

 dwindling away of this trunk. But the same material affords an equally clear example of 

 a change in the trend of evolution; for the former tendency toward the development of a 

 "trunk" has been arrested and a rapid secondary shortening of the bony tract between the 



