Develojjmejit uf the Food-Fishes. 451 



invest the yolk, and on tlie second day usually covers more 

 than a third of its surface. Early on the third day tlie 

 equator is passed, and at the fifty-fourth or sixtieth hour tlie 

 blastoderm generally envelops two thirds of the yolk. The 

 cephalic end of the embryo remains stationary in the forms 

 under consideration ; but as the caudal extremity keeps pace 

 with the advancing periphery of the blastoderm, its increase 

 in length must take place in the region of the trunk. The 

 blastopore reaches its maximum at the equator of the ovum^ 

 and when that is passed its circumference continuously de- 

 creases. The rim does not increase appreciably in density or 

 breadth, but, on the contrary, its substance diminishes, and 

 this must be so if, as it progresses, it contributes to the 

 investment of the yolk. When the blastopore has so far 

 decreased as to appear merely as a minute aperture (on the 

 fifth or sixth day) at the posterior extremity of the embryo, 

 the rim is recognizable only as an aggregation of cells — the 

 cells of the coalesced margin. This remnant is probably 

 used in the formation of the anal section of the mesenteron 

 and other structures ; but it does not appear that the caudal 

 plate is formed directly and almost solely, as Ryder main- 

 tains, out of these cells, the tail, like the rest of the trunk, 

 increasing in length by the addition of mesoblast somites. 

 Still more questionable is the theory of Rauber and His, 

 adopted by Ryder, that the hind portion of the embryonic 

 trunk exemplifies the phenomenon of concrescence, since it is 

 not supported by study of the living embryo, and sections 

 reveal no trace of a median fissure or line of apposition con- 

 tinuous with the longitudinal vertical plane of the anterior 

 region. Indeed the caudal plane of symmetry is at right 

 angles to the plane of symmetry in the rest of the trunk, for 

 the tail lies sidewise upon the yolk, and apparently develops 

 and continues in a state of torsion until the embryo is free. 

 Further, the solid condition of this portion of the embryo is 

 maintained till a comparatively late stage, when the medul- 

 lary canal finally penetrates it, as the first median fissure 

 which divides its cells. 



The complete diiTerentiation of the notochord coincides in 

 many species with the closure of the blastopore. This is the 

 case with Gadus merlangiiSj G. morrhua^ P. limanda. and 

 P.flesiis-^ but in the case of G.oiglejinus and T. gurnardus 

 the closure of the blastopore is one or two days later. 



Nothing noteworthy was observed respecting Kupfer's 

 vesicle ; • it has the same structure in the various species 

 though in T. gurnardus (in which it was not observed until 

 twenty-four hours or more after the blastopore had closed) it 



