EMBRYOLOGY OF THE SEA BASS. 
221 
trifugal delamiaatiou also takes iilace iu the diiectioii of the dotted line. Now, by a 
I slight alteration in tlie course of the lines of growth, the zone iu which they all meet, 
I or the apical line, may be made to moYe centrifugally or centripetally, and if, for 
instance, it take the position b, an extremely wide randwulst will be established. In 
this hypothetical randwulst, as in the real one, there would at lirst be no differentiation 
of layers, but by the quick and simple process of delamination the under cells of 
the wulst could be conv^erted into a separate layer, as actually occurs in the real 
' wulst. By moving the apical line far enough towards the center, the entire germ 
ring could be established by delamination, without there being any need of an actual 
[ ingrowth. 
The germ ring as such reaches the height of its development about 20 hours after 
, fertilization (Figs. 31 and 35, PI. xcii. Fig. 41, PI. xciii. Pig. 48, PI. xoiv, all of the 
: same age). The thinning out of the central region to form the subgermiual cavity 
I (s. y. c.), which was begun in earlier stages, has been continued. The thin region 
now, however, has very detinitely circumscribed bounds. It is inclosed on all sides by 
the germ ring (Fig. 34) in the region of which, especially in the neighborhood of tlie 
embryonic pole, the upper layer of cells (which may now bo called ectoderm) remains 
J thick (Fig. 44). The thin region, or extra-embryonic part of the blastoderm, consists 
[ at this stage of three layers of cells — the epidermic stratum and two strata of nervous 
j layer cells — which are still plump polygonal bodies. This region continues to grow 
'! thinner, but reaches its ultimate condition before the blastoderm has covered the yolk. 
It is then a thin membrane, made up of epidermic pavement cells and one or two layers 
of somewhat flattened cells. 
Historical . — To Gotte belongs the credit of the discovery that iu the Teleost 
embryo there is a process of invagination leading to the formation of the primitive 
hypoblast, which iu its turn splits up into entoderm and mesoderm. The account given 
' in his first communication (1809, 15) was scarcely an exact description of the facts, but 
his second pai^er (14) contains a more accurate and detailed account than any I have 
found iu subsequent papers. 
In his first communication Gotte described the edge of the blastodisc as suffering 
an actual involution to form the under layer, and this account was repeated by Haeckel 
' in the description he gives of Teleost development iu the “ Gastrtea Theorie” (19). In 
j 1873 Oellacher published his well-known paper on the Trout (33), in which he claimed 
there was no invagination at all, but that the eccentric thinuiug out which led to the 
formation of the subgerminal cavity also left one portion of the peripheral region much 
thicker than the rest (embryonic anlage). This thicker portion of the periphery then 
split into the ectoderm, mesoderm, and entoderm of the embryo (delamination theory). 
Shortly after the appearance of Oellacher’s paper Gotte published his second account, 
l! with which my own description agrees in all essential particulars. The one point of 
j difference concerns the randwulst. What Gotte has described and figured under this 
name is really a stage of the germ ring, in which the free ingrowth has already begun. 
It corresponds with the ring shown in my Fig. 42, PI. xcili, or more highly magnified 
in Fig. 40, PI. xciv (anterior pole), and Fig. 41, PI. xciii, right half (posterior pole). 
; As will be seen from the following brief abstract of Gotte’s description, the only 
' improvement I have been able to make upon it lies in having defined the randwulst 
with somewhat greater iirecision. According to Gotte the thinning out by which the 
I subgerminal cavity is established is the result of the direction of cell growth, which 
