igoy] Obscure Generalisations 



cheek — rock cod, sculpins, flatheads, gurnards, sea 

 poachers, and snailfish — large fins, rough scales or 

 bony plates, and strong spines rise by degrees, then 

 give way progressively to feeble fins, naked skin, and 

 obsolete armature. The other series — butterfly 

 fishes, moorish idols, surgeon fishes, trigger fishes, 

 leatherjackets, puffers, porcupine fishes, and head- 

 fishes — lose their spines and pennant-like fins, their 

 scales and teeth, being in the end reduced to a great 

 head close behind which is attached the caudal fin. 

 My only explanation of these phenomena is that 

 specialization was overdone and thus became a posi- 

 tive hindrance. If this view be correct it certainly 

 is not a matter of orthogenesis as conceived by 

 Eimer, for that he interprets as the result of an im- 

 pulse from within. 



A second anomaly in the development of fishes may taw of 

 be stated as the Law of Vertehrce. Both Gunther and ''""^'^"^ 

 Gill observed that in two different groups — flounders 

 and wrasses — the northern species have more verte- 

 brae than those living in the tropics. This fact, sup- 

 posed by my two colleagues to be mere coincidence, 

 I followed much farther, finding it to be an almost 

 universal rule that to north or south of the tropics 

 vertebrae become smaller and more numerous, so that 

 around the Arctic Circle the usual number is fifty or 

 more, while in equatorial waters it is only twenty- 

 four in most groups and falls to eighteen in some cases. 

 In river and deep-sea fishes it is also more or less in- 

 creased. 



In my several papers on "Latitude and Vertebrae" 

 I have elucidated the facts, and show that natural 

 selection has a bearing on them, though fundamental 

 causes still remain obscure. It would, however, 



C 249 3 



