DIFFERENCES IN HISTOLOGICAL STRUCTURE OF TEETH. 465 



For the deutines of Phycis aud Motella (the rockling), figs. 

 15 and 16, which fish do not stand far apart from the ling 

 and hake, have the vascularity so far reduced that its pattern can 

 hardly be positively deciphered, though, so far as any evidence 

 does exist, it appears to resemble that of the genus Gadus more 

 closely than that of the ling. 



It is, of course, fully recognised that no linear arrangement 

 of genera and species can ever truly represent relationships, 

 but still the classification by tooth structure departs somewhat 

 more widely from that adopted upon general grounds for the 

 deviation to be thus accounted for; whilst if we were to take 

 into account simply the quantity of vascular canals present we 

 should have still more discrepant results, for it would then 

 come about that some species of the genus Gadus would be 

 interspersed amongst other widely different genera, as is here 

 shown. 



2. Merlucius ) 



7. Molva / ^^'^^^^ system of canals). 



1. Gadus morrhua. 

 9. Raniceps. 

 10. Brosmius. 

 l.Gadusluscus. 

 1. Gadus mi nutus. 

 1. Gadus poUachius. 

 1. Gadus aegle finus. 

 6. Lota. 

 1. Gadus virens. 



4. Uraleptus. 



5, Phycis. 



8. Motella. 



3. Lotella (no vascular canals at all). 



It is true that the genus Gadus has been sometimes divided 

 into several genera, but even in that case they are very closely 

 allied genera, and cannot possibly be regarded as interspersed 

 amongst other genera with any propriety, so that the conclusion 

 is irresistible that these differences of tooth structure only to a 

 limited extent follow the lines of the general affinities of the 



