()1S Mit. It. II. wiiii-Kiior.-^p; ()\ TiiK [Ai»i-. 5, 



OlMilDiUM h.Mlli.VTiM. (J U'^iilaic.-^, Oiiliiiliida-.) (Plate L. 

 iig. 3U.) 



All the iiH'dian fins of this fish are continuous with one another, 

 and the caudal proper may in this particular case be limited to those 

 (lermotrichia whicii are not supported by the typical i-<idial or 

 somactid of the dorsal and anal fins. This confines the caudal 

 rays to those eight or nine supported by a large terminal hypural, 

 often deeply and irregularly cleft, which is firmly coalesced with 

 the last vertebral segmejit ; the urostyle therefore is reduced to 

 the merest vestige. A diminished neural ai'ch is to be seen on 

 the last centrum, and the nerve-cord runs along the whole length 

 of the dorsal edge of the hypural. Thus the caudal fin itself is 

 exceedingly simple in foi'm and very highly specialized. The 

 neural and luemal arches of the penultimate vertebra have a 

 eartilaginous shaft fused to their posterior side ; in reality, thei-e- 

 fore, they may be considered epural and hypui'al respective!}', 

 l)ut bearing no fin-rays. 



General remarks on the Acanthopteryylan Caudal Fin. 



A sufiicient number of types from various divisions of the 

 Acanthopterygii have been dealt with to show how, in a single 

 sub-order of Teleosts, the structure of the tail-fin n\ay vary. 

 Even in the same division of the sub-order, vastly difierent types 

 are found ; for example, in the Perciformes, the geneia Serranns, 

 Box, Cantharus, and others have features characteristic of the 

 less specialized homocercal tail, such as a long and free urostyle 

 and a lai-ge number of hypiual bones attached to the last vertebral 

 .segment ; on the other hand, the Wrasses (Labridai) have lost 

 these lowly specialized features and attained to the rank of liigh 

 si)ecialization. The Scleroparei provide a similar contrast, 

 ,Scorpa')ia having a much less advanced type than Tr'ujla. It 

 would be difficult to account for such variations between types 

 otinn'wise so closely related, except on phy.siological grounds, but 

 it indicates that the tail-fin is not a stife guide, or ratln-r, a 

 wholly satisfactory character, by which fishes can be cla.ssified. 



In the main, however, advanced homocercy is exliibite(l among 

 Acjinthopterygian caudals, except for the Perciformes ; this 

 division possesses the greatest nund)er of less specialized tails 

 (the Labrida; are the only family with much exception), but it is, 

 as a division, the least specialized among the Acanthopterygii. 

 From the foregoing descriptions it will be seen that the Jugulaies 

 possess the most advanced types of homocercy ; of the six geiun-a 

 which have been sul)jectcd to examination, viz. Cravoscojms, 

 Trachinus, Lej)ado(/asler, Blennius, Crisiicrps, and Op/ridiuvi, all 

 po.sse.ss advanced features in the caudal liu ; liut (he Jugulares are, 

 as a division, among the mo.st specialized of the Acanthopterygii. 

 The Zi'orhond)i occupy an intermediate position, and all mend)ers 

 have Acrv .similar tails: they have a greatly ri-dnced urostyle 

 fused with a livpiu:d, and al.sn a lai'gc nuiubcr of lixpurals to the 



