BULLETIN OF THE UNITED STATES FISH COMMISSION. 289 



usual mode. Sucli unusual acceleration or retardation in the develop- 

 ment of certain structures in various species of teleosts is not unusual, 

 and wonid be as likely to aftect the segmentation of the mesoblastic 

 blastodermic rim into muscular segments as any other part of the em- 

 bryo. There is another most serious objection to the unqualified accepta- 

 tion of Balfour's theory of the growth of the embryo from the edge of 

 the blastoderm without further addition from that source. If we do not 

 admit that the blastodermic rim becomes transformed into the body of 

 the embryo, what becomes of it? Nothing can be more certain than 

 that, ui)on its closure, little or nothing is left of it; it has apparently 

 been incori)orated into the embryo's body. 



This view appears to be well sustained by what may be observed in 

 the development of Belone. Up to the time when the embryo may be 

 said to be fairly outliued as in Fig. 6, forty-three hours and forty min- 

 utes after impregnation, the material of the blastoderm and embryo has 

 acquired little or no increase of bulk in consequence of the incorporation 

 of portions of the massive yelk. In Fig. 6 we see that the embryonic 

 body occupies about a quarter of the circumference of the yelk. The 

 blastoderm has grown down over and inclosed more than half the yelk 

 globe, and its rim is contracting at the tail to comi)lete the closur6. 

 When this is accomplished, the point where the closure takes place en- 

 tirely diappears; the edges of the rim have been so perfectly fused 

 together that the point of union, marked at first by a pore behind the 

 end of the tail, with radiating wrinkles running out from it fifty-one 

 hours after development began, as shown in Figs. 8 and 10, has soon 

 after completely vanished. The material of the slowly contracting rim 

 is finally fused into a solid flat plate of cells at the caudal end of the 

 embryo, after the membranes of the latter — epiblast and hyi^oblast — 

 have inclosed the yelk. The conversion of this caudal plate into the 

 mesoblastic, epiblastic, and hyi)oblastic structures of the tail end of the 

 embryo accordingly appears to me to be beyond question. But I would 

 not commit mvself to an adherence to the doctrine that the embryonic 

 body was formed by a gradual coalescence of the thickened edge of the 

 blastoderm from before backwards along the median line. If the reader 

 will observe Fig. 9 he will see that the annular blastodermic rim r, as 

 it approximates the closed condition at the fifty-first hour, is not circu- 

 lar, as in Fig. 5, but decidedly oval. The sides of the oval blastodermic 

 annulus are now approximated more rapidly than the ends, as we see 

 still further exemplified in the oval pore like openings in Figs. 8 and 10. 

 It is, therefore, probably nearest to the truth to say that the embryo 

 grows in length both by intussusception from behind forwards of the 

 blastodermic rim as well as by the coalescence of the latter, not along 

 the median line, but by a gradual fnsion as it is finallv closed over the 

 yelk. 



The segmentation of the mesoblast proceeds in the usual way in 

 Belone from before backwards, as shown in Fig. G at so, and there is no 

 Bull. U. S. F. C, 81 19 iriay 1 9, 1 8 8S, 



