[31] EMBRYOGRAPHY OF OSSEOUS PISHES. 485 



liug there was uo trace of the presence of blood corpuscles, I can con- 

 firm. The same is true of Crenilabrus, JuUs, Scorpa'ua, Fierasfer, etc." 

 In that the yelk hypoblast contains free nuclei in what appears to 

 be a continuous and homogeneous sheet of superficial protoplasm, with- 

 out any evidence of division into cells or segments, it is truly of the 

 nature of a syncytium as defined by Haeckel, or it may be defined as a 

 multinucleated protoplasmic layer. A. Rauber also speaks of it as a 

 I)lasmodium, a term borrowed from cryptogamic botany, and first ap- 

 plied to that most remarkable of substances to be met with under rotten 

 wood and damp leaves in moist glens and representing a stage of de- 

 velopment of a very singular order of Fungi. Plasmodium is as near the 

 ideally structureless non-nucleated condition as one of the Monera, but 

 in the case of the germinal layer of the egg of osseous fishes, while it is 

 apparently devoid of nuclear bodies up to the time of the first cleavage, 

 after that it acquires them and really becomes a syncytium, as before 

 stated. The point insisted upon by Hoffman is of great weight in rela- 

 tion to the part it plays in the development of the blood, and in the 

 later stages of embryonic development we shall find that the blood 

 actually develops from it and that tlie larval blood-vascular system 

 of the yelk is in part actually channeled out of it su])erficially. This 

 idea was first tacitly formulated by Professor Carl Vogt, now of Geneva, 

 in 1842, when he made use of the term couche JuvmatogSne in describing 

 the development of Coregonus jjala'aj where it could scarcely escape ob- 

 servation if the lowermost germinal layer in that form shares in the 

 <leveloi)mentof the blood as conspicuously as in the embryos of our own 

 whitefish, Coregomis albus. 



7. — Segmentation of the germinal disk. 



When the protoplasmic streams over the surface of the yelk have 

 carried the principal portion of the germinal protoplasm to the disk, 

 these disappear and the cod's Qgg no longer presents the appearance 

 shown in Fig. 7. When the aggregation of the disk is completed it 

 presents a discoidal, biscuit like form, with the edges blunted or rounded 

 off" and thickest in the center. While all evidence of the strands of 

 germinal matter radiating from the edge of the disk has disappeared, a 

 thin veil of germinal protoplasm still remains behind to cover and in- 

 clude the yelk sphere. This stratum in the cod's egg is exceedingly 

 thin and is continuous with the germinal protoplasm of the disk all 

 round the margin of the latter previous to the advent of segmentation. 

 The same condition appears to hold in the case of Salmonoids, Olupe- 

 oids, and Cyprinodonts, but in all of these the layer of germinal matter 

 left behind as an envelope for the yelk seems proportionally thicker. 



After the germinal disk has been fully developed and has assumed 

 the biscuit form already alluded to, at about the sixth hour, in eggs 

 which hatch in twenty days, it begins to elongate, becoming wider in 

 one direction than in the other, and at the same tinie more depressed, 



