[63] EMBRYOGRAPHY OF OSSEOUS FISHES. 517 



the dorsal and ventral median line. With the progress of development 

 it becomes more conspicuous, growing in height, so as to soon be very 

 much wider, as may be seen in Fig. 34. Its first appearance is heralded 

 by a faint doubling of the skin upon itself, so as to project outwards as 

 a median ridge, extending from the point of origin of the tail at its ven- 

 tral margin from the yelk-sack back over the caudal extremity and pro- 

 gressively forward over the median dorsal line towards the head. By 

 the time the embryo leaves the egg this fold extends forward on the 

 back as far as the pineal gland, or to a point just behind the forebrain, 

 as may be seen in Fig. 40. Its development, however, is continued even 

 somewhat farther forward fourteen days after hatching, as shown in 

 Fig. 45, until it ends almost immediately between the nasal pits. At 

 this time its extreme anterior extent gives to the young cod a very sin- 

 gular appearance as viewed from the side, such as is not met within any 

 other form which I have studied. In outline, as viewed from the side, 

 the young fish now bears a resemblance to the conventional representa- 

 tions of the dolphin i^ old sculptures. 



The natatory fold is now actually wider than the caudal portion of 

 the trunk, but it is quite thin and comprises only the skin folded upon 

 itself, its whole thickness being mainly, if not entirely, derived from 

 the epiblast. At first, in all forms known to me, the caudal portion of 

 the natatory fold is rounded in outline, as seen from the side, but may 

 assume a fan-shape, even before a single caudal fin-ray has been devel- 

 oped in it, as is the case in Alosa and Pomolobus. In others, again, the 

 rays begin to develop before the caudal portion of the primitive median 

 natatory fold has become fan-shaped, as may be seen in Salmo and 

 OncorhyncJms. In still others there is no continuous median fin-fold 

 developed at all, as in Gamhusia^ Siphostoma, and Eippocampus, and the 

 median fins grow out at first as short, local, dermal folds, in which fin- 

 rays soon afterwards develoj). In those forms in which the unpaired 

 fins are developed from a continuous median fold, the dorsals, anal, and 

 caudal are evolved only in certain regions of the fold itself, the por- 

 tions of the latter, which do not become fin rudiments, atrophy. Balfour 

 says (Comp. Embryol. II, '63) that "the dorsal and anal tins are devel- 

 oped from this fold by local hypertrophy." The process, however, when 

 narrowly studied, presents features the significance of which cannot be 

 fully apprehended under the term hypertrophy. As stated at the out- 

 set, the median larval fin-fold is at first a mere outward duplication of 

 the skin containing no raesoblastic tissue between its laminse, but as 

 soon as the positions of the fin-rudiments are defined we may note that 

 there has been an outgrowth of mesoblastic tissue into these regions, 

 causing them to become thicker and less transparent. With the prog- 

 ress of this process, the raesoblastic tissue gradually advances toward 

 the margin of the fohl, insinuating itself between the epiblastic walls of 

 the continuous fin of the larva. Soon afterwards it becomes evident 

 that the fin-rays are becoming differentiated by the mesoblastic tissue 



