984 i:i:r()KT of commissioner of fish and fisherif.s. [4] 



tra, are (iscil in the sense in whieli they were tirst nsied by Jhixk\v in 

 1859 in his discnssion of the development of the tail of Gasterosteus. 



The word opisthnral applies to apophysial elements in contact with 

 the nndei- side of the urostyle, bnt separated from the liy])ural pieces 

 by an interspace, as in Aminrus. 



The wh(jie of tiie b<»dy of tishes from the vent backwards to the end 

 f of the tail will be fretpiently spok(Mi of as the uronome. The exserted. 

 or rather the dei;eneiated, portion bi'hind the nrosome, as fonud in Sty- 

 le2)horns^ Ghimaro monxtro.su. and the larva of Lepidosteus. will be called 

 the opinthurt', becanse it is behind the trne or secondary candal, which 

 is developed aft<n- the former has begun to lose its fnnction throngh the 

 substitution of the latter. The word opiaihure is by no means synony- 

 mous with nrosttfle, because it includes more, as in some of tlie cases 

 mentioned abo\»'. It (•omi)rises tiie end of the chorda around which 

 the osseous urostyle is develoi>ed : bands of muscalar tissue on either 

 side, which are fretpu'utly rudimentary or im])erfectly formed myo- 

 tomes or muscular srgments, besidt* connective tissue, the whole being 

 covered exttniially bv inU'guiuent, which is continuous witli that over 

 the rest of the body. 



In the case of a perfectly lophocercal form an imaginary horizontal 

 plane cutting longitudinally through the center of the iiotochord will 

 divide the urosome of a larval lish into an inferior and a superior half, 

 and these halves may. roughly speaking, be said to be the dorsal and 

 ventral counteii)arts of each other. All of the skeletal elements above 

 the notochord may therefore be saiil to be epaxkil, those below it hy^)- 

 axlnJ in position. This also appli«^s to the median skeletogenous tract, 

 the i)arts of which may ho spoken of as epaxial and hyjjaxial in posi- 

 tion. 



The skeleiitgenous txats of tish embryos seem to be largely intermus- 

 cular or intersegmental. i>erichordal, hypaxial, and epaxial, so that it is 

 l)ossible to trace the segmental development of the median and paired 

 tins in early stages, and leads us to sus]>ect that sucdi a segmental ar- 

 rangement may be ettected from such Joci^ and thus determine the rela- 

 tions and serial arrangensent o\' the rays and basal ]>ieces of all of the 

 tins, \\liile the development of the lays and their lu'oximal supports 

 is often obseiired by jn'ocesses of coalescence, degeneration, and shift- 

 ing, due to grow (li. future ontogenetic studies will unquestionably place 

 the tlieory of all of tbe tins uj>on souu' such solid basis of observed fact. 

 When the skehtoiienons tract is s]>oken of heie it is used with such im- 

 itlications. 



The einbiyoniv skeleton is ;it fust mainly cartilaginous, and around or 

 Ml this cartilage — usually tlie former — ossific deposits are laid dovrn in 

 a i)et!uliar homogeneous kind of membrane which invests the cartilages 

 [Ttltostd). This sort of a homoifincous memttraue may appear in the 

 s!<eletogenous tracts of detinite regions nearly as .""lOon as cartilage, as 

 in the case of the nuMiibranous rndiments «)f the shoulder girdle of 

 Teleosts. 



