[17] EMBRYOGRAPHY OF OSSEOUS FISHES. 471 



been uiauifested. With this accumulation of the germiual matter the 

 correspoudiug pole of the egg also becomes heavier. 



In its singular progressive movement the germinal layer is observed 

 to become thicker at the lower than at the upper pole of the egg, and 

 to bulge upwards into convex elevations on its inner surface next the 

 yelk. This migration proceeds until the yelk at the upper pole is 

 almost exposed, as in Fig. G, which shows the relation of the germinal 

 matter to the yelk one and a half hours after impregnation. While there 

 are no oil spheres present to buoy up the eggs, as in the mackerel and 

 moon-fish, the specific gravity of the germinal matter is greater than 

 that of the yelk, so that it always assumes a position on the lower side 

 of the yelk. Should the egg be turned round so as to bring the disk 

 uppermost, the yelk will be gradually turned by the gravity of the disk 

 until the latter regains its nethermost position. During its migration 

 the germinal protoplasm eventually arranges itself in radiating bands, 

 which sometimes anastomose, and all trend towards and join the edge 

 of the incipient disk below. Later these bands develop nodes or en- 

 largements, ^^rn, along their courses, as shown in Fig. 7, and pour their 

 substance into the disk, which is now defined three hours and forty min- 

 utes after impregnation. If a granule in one of these bands is watched 

 for a time it will be noticed that it exhibits a more or less decided pro- 

 gressive movement. 



At the time the disk is defined its inner surface at first presents irreg- 

 ular rounded elevations, which gradually subside, when the under side 

 of the disk becomes flat. Then the outer surface of the disk is elevated 

 into one or more large rounded, prominences, which in like manner even- 

 tually disappear. These are some of the amcebal phenomena already 

 alluded to. 



It is very important for us to make a distinction here between the 

 mode of formation of the germinal disk of the cod and that of the Clu- 

 peoids, as worked out by Kupft'er and myself, and of Tinea, as des(!ribed 

 by Van Bambeke. It would appear that we concur in the opinion that 

 in the latter more or less protoplasm destined for the development of the 

 germinal disk is derived from the center of the vitellus, into which, as 

 may be seen in the ova of Alosa, fjlupea, and Pomolobus, the external 

 germinal layer sends processes which in the sliad and branch herring 

 look like hyaline roots passing down amongst the yelk spheres. At 

 a later stage of development this arrangement seems to disappear, 

 and the yelk spheres, although still evidently involved in a meshwork 

 of germiual matter, do not have the same amount of the latter insin 

 uated between them as at first in the vicinity of the develo]jiug ger- 

 miual disk. On the other hand, in the eggs of Gadus, TylosuruSj Cy- 

 bimn, Elacate, and ParepMppus, the germinal matter forms a distinct 

 outer coating over the vitellus, and these forms do not have any dis- 

 tinctly marked yelk spheres, as in AJosa, involved in germinal proto- 

 plasm, except in the case of Elacate, where the yelk spheres are very 



