Teleosts and Birds 



299 



initiation of nuclear contribution to periblast 

 formation, and the appearance of special 

 osmiophile and other staining properties has 

 been reported in the peripheral cells destined 

 to form the germ ring (Rauber, 1883; 



Fig. 109. Results of removal of blastodiscs with 

 varying amoiuits of yolk, follov^^ed by culture in 

 Holtfreter's solution. Interrupted arrows: work of 

 Oppenhetmer ('36c) on Fundulus; solid arrows, 

 that of Tung, Chang and Tung ('45) on Ca,rassius. 

 In the first series, the blastodiscs were invariably 

 removed clean of yolk; in the second, a small 

 amoimt of yolk was always left adherent, and in 

 early stages differing amounts of yolk, as indicated, 

 were purposely included. The diagram shows only 

 the most radical results; fragments capable of form- 

 ing complete embryos may also develop incom- 

 pletely in many cases. 



Agassiz and Whitman, 1884) in Gobius and 

 Ctenolabrus; there is a possibility that an 

 important step in differentiation is involved 

 here. 



Morgan was the first to report that the 

 entire yolk mass is not necessary to the dif- 

 ferentiation of the embryonic axis in teleosts. 

 Figure 109 summarizes and compares the 

 relations worked out more recently in Fun- 

 dulus and Carassius, independently. Blasto- 

 discs were removed, with or without some 

 yolk, as indicated, and cultured in saline 

 medium. Results ranged from non-gastru- 

 lating "hyperblastulae" through varying de- 

 grees of abnormal or defective embryo for- 

 mation, to quite perfect embryos. In general, 

 the results suggest comparable progressive 



changes in the two forms. Fundulus blasto- 

 discs may form hyperblastulae after isolation 

 at any stage. The 32-celled stage is the 

 youngest isolate able to form an embryo. In 

 Carassius, the eight-celled blastodisc, iso- 

 lated except for a minimal quantity of yolk, 

 is the oldest stage capable of forming a 

 hyperblastula. Occasionally this type of iso- 

 late may form an embryo. In younger stages, 

 three-fifths of the yolk must be included 



Fig. 110. Summary of results of Tung and Tung 

 ('44) on raising both halves of bisected eggs. The 

 frequency and types of result are schematized be- 

 low. A half-egg may form a complete or nearly 

 complete embryo, a defective embryo, or a hyper- 

 blastula. Out of nine successfully raised pairs, 

 three consisted of two nearly complete embryos, 

 one of two hyperblastulae, and the rest, as indicated, 

 showed intermediate combinations. 



(1- and 2-celled stages) or half the yolk 

 (4-celled stage) if an embryo is to result. 

 Explants with less yolk fail to gastrulate 

 or to form axes. These results may possibly 

 indicate, as the authors suggest, a flow of 



