610 Mn li. ir. \viuT;:ir(iisK ox tiii: TApr. T), 



Orlior rpferonci'S luii^ht iilso ho (|U()te(l, Imt tlicsp will he 

 sulHcient to show what uiiseonceptions sue held. J)e;iliiig with 

 Ciiniiiii^'liam's icinark.s first, a lefereiiff to either of the two 

 figures given will show at a glance that the tail is not " perman- 

 ently diphycercal " and is not " coniposed of dorsal and ventral 

 rays equal in nuniher." In all, eight different species have been 

 very carefully examined, and any of the following remarks ajjply 

 equally to all ; thei-e are always more hyjiaxial fin*rays than 

 opaxial ; this is due to the fact that a broadly expanded hypural 

 bone is attached to the last centrum. Thus the tail is a highly 

 specialized homocercal form. 



Tliis author has apparently never been fortunate enough to 

 secure young forms of Gadida^ for detailed examination of the 

 caudal fin structure or he would never have sai<l tliat dui-iiig the 

 devi'lopment the heterocercal stage is "entirely wanting." 



G'adas viimUus of one inch length shows an upwardly directed 

 urostyle along the dorsal bi.\^g of the last hypural as clear]}- as in 

 the flat-fish ; this becomes much reduced in the adult and is 

 in fact almost completely lost ; and since the extremity of tlie 

 vertebi-al axis is almost precisely the same in all adult Cadidje, 

 it is quite reasonable to suppose that the young foi'ms in every 

 ca.se are similar to G. iiiinutvs. Cunningham's error in this 

 connection led him to sup])ose that the Gadida; "are descended 

 from Crossopterygian Ganoids with diphycercal tails and have 

 never passed tbrough a heterocercal condition." The latter 

 statement, however, falls through. 



A.ssuming Cunningham to be correct as regards the diphycercal 

 ccmdition, Uoulenger comes to the conclusion that the Gadida' are 

 descended from a homoceival form, such as the ]^eryci«ho, but 

 that homoceiry was entirely lo.st, and the " isocercal (?. e. the 

 diphycercal) condition .... is the result of the formation of a 

 new caudal fin." But development proceeds just as in any highly 

 specialized homocei'cal form, and hence the caudal fin of the 

 Gadidfe cannot be considered a new thing. 



The caudal fin of this family further conforms to adv.iiu'cd 

 honiocercy in the gre.at reduction of the neural arch of the 

 ]icunltiniate vertebra ; in all previous figures of the Gadoid tail, 

 the last vertebra but one is shown bearing an epural bone ; but 

 in all the Gadida' examined, this has pi'ovcd not to be the case; 

 immediately dorsal to the penultimate vertebi-a there are alway.s 

 two dors-al caudal ladials perfectly frei' and never fused with any 

 f)thcr jjart of the neighbouring skeleton, Hypaxially, the same 

 vertebra always bears a hypuial bone. 



Thus the caudal fin of the Gadida' is vmdenialdv a specialized 

 form of homocercy. 



A f A XT n (I 1' r K u v c i r. 



A preliminary noti- will be necessai'v before entering into 

 details of the types chosen in this sub-order. In many foiiiis 

 in which a free uiostyle is present, there is a fin-ray .'iuj port 



