568 KErORT OF COMMISSIONER OF FISH AND FISHERIES. [114] 



tlieory, but tends rather to become narrower vertically as it advances 

 over tlio yelk globe toward the level of the line 2, Fig-, G, while it is 

 gradually coiisti'icted in diameter as it grows i)ast the level of lines 

 3 and 4, beyond which it closes at x. In ova with a very large yelU, 

 like those of the salmon and silver-gar, the blastoderm does not close at a 

 point opposite to that where the germ was developed. This has given rise 

 to some dispute amongst embryologists, a disagreement which may be 

 explained by the diagrams A and G. If we suppose A to represent a 

 very large yelked osseous fish ovum, Avith a as its germinal pole, the 

 body of the embryo will grow, say, to the point A, the blastopore closing 

 at I. The rim will then cease to advance at the tail of the embryo or 

 at the point x in Fig. G, when, as shown in the latter, the portion of the 

 rim on the opposite side of the agg will liave to advance faster from 

 the point ii onward towards ill, iv and v, closing at x instead of x*. 

 When the agg is a medium sized one, like that of Alosa in Fig. B, the 

 embryo stojjs growing in length only when the tail reaches the oppo- 

 site pole, as shown in Fig. A; if the egg is still smaller, the embryonic 

 axis may continue to grow beyond the opposite pole, so that the blas- 

 toderm does not close till it reaches the point marked iii beyond which 

 the tail buds out at C. This third form also requires another mode of 

 closure of the blastoderm, differing from the two preceding types. 



In a large ovum, according to Fig. A, the embryo ceases to grow in 

 length when it has extended itself over aii arc of the yelk globe of, say, 

 125°; in Alosa, or the second form, it extends its growth throngh an 

 arc of 180°; in the third and smaller type of ovum {Garassius), it may 

 grow to a greater length and embrace an arc of 230° on the surface of 

 the yelk sphere. In the first type the rim of the blastoderm is some- 

 times drawn out into an oval prior to closing, as shown in Fig. C taken 

 from Tylosurus. These different modes of growth in length of the bod- 

 ies of the embryos of different species of fishes are matters of oK«erva- 

 tiou with me and go far toward reconciling the differences of opinion 

 which have been expressed by different observers as to the growth of 

 the blastoderm over the yelk. It is evident, at any rate, that what may 

 be observed on this point in one type may not apply to another. 



The segmentation cavity sc, Fig. G, extends laterally, with the advance 

 of the rim of the blastoderjn, towards the opposite pole of the Q.gg, and 

 does not disappear, as held by Haeckel and Balfour. In Fig. E it ex- 

 tends between the yelk hypoblast and its epiblastic outer coverings 

 from a clear to the tail cm, and from one side of the body down, over, 

 and around the yelk to the other side. It may be seen devcloi)ed to a 

 remarkable extent in some forms, as in Ci/bium, Corcgonus, and Alosa. 

 Figs. 47 and 48 show it in two stages, under the blastoderm of the cod. 



In Goregonus, tlie oil-drops, by their buoyancy, bulge the yelk hyi)o- 

 blast upwards into the cavity, so that innnediately over each consid- 

 erable droplet there is a ijerceptible rounded elevation of its floor. 



The various cavities which different observers claim to have seen dur- 



