PRIMITIVE STKKAK 



35 



extra-embryonic ccelom is also occupied by delicate strands of the same tissue 

 stretching between the yoik-sac and wall of the blastocyst. These constitute what 

 has been named the magma reticulare. 



While this loose mesoderm is developing, the entoderm, forming the roof of the 

 entodermic sac, becomes in Tarsius thickened into a many-celled plate (fig. 51). This 

 plate is produced by proliferation from the entoderm (Hubrecht), and is continuous 

 at the margins of the disc with the yolk-sac mesoderm, in the formation of which it 

 seems to share. It is therefore a second source of mesoderm- cells, quite independent 

 of the earlier one. It is named by Hubrecht the protochordal plate, but it will be 

 here referred to as the primitive entodermic plate, to avoid any theoretical 

 implication. 



In Selenka's figure of a blastoderm of Semnopithecus nasicus (fig. 54) the mesoderm is repre- 

 sented at this stage as extending forwards into the disc from its hinder end. It would therefore 

 seem to be derived from the same source as 

 the mesoderm of the connecting stalk, which, 

 however, as already said, he refers to the ento- 

 derm. In Peters' and in Graf v. Spee's early 

 human ova a similar layer is seen (fig. 55). 



From this description it will be gathered 

 that the ectoderm and entoderm are everywhere 

 separated in the Primates by a middle layer 

 before there is any sign of a primitive streak. 

 Although different from the conventional ac- 

 count of the origin of the mesoderm, there is no 

 doubt that the facts are as stated. Their 

 theoretical significance will be dealt with in a 

 later paragraph. 



Up to the stage now reached, only 

 that part of the three-layered blastoderm 

 which we may call the head-plate, because 

 it will form the extreme head end of the 

 embryo, is laid down, and we have next 

 to describe a series of stages by which the 

 embryonic axis, forming the trunk, is 

 developed. As in the typical case de- 

 scribed above, the germinal disc enlarges 

 (fig. 56), and Hensen's knot (protochordal 

 knot, Hubrecht) appears as a thickening 

 of the ectoderm. This thickening extends 

 inwards and forwards between the two 

 primary layers (fig. 57), and is continuous 

 with the thickened entodermal plate in 

 front. It then becomes fused with the 

 entoderm on its under aspect. The primi- 

 tive streak next appears as an extension 

 backwards of the ectodermic thickening. 



that from its sides, as well as from Hensen's knot, wing-like masses of mesoderm 

 extend laterally between ectoderm and entoderm (fig. 58). They are formed from 

 cells budded off from the streak, and from this period onwards the new mesoderm 

 of that part of the blastoderm which lies between the head-plate in front and the 

 connecting stalk behind, and which gradually increases in length as the embryonal 

 axis is laid down, may be considered as arising from this source. 



Hubrecht describes in Tarsius a tract of middle-layer cells springing from the 

 entoderm over a ring-shaped area, continuous in front with the primitive middle- 



D2 



V > 



FIG. 51. MEDIAN LONGITUDINAL SECTION 

 THROUGH THE EMBRYONIC PLATE AND YOLK- 

 SAC OF TAKSIUS AT THE SAME STAGE AS 



IN FIG. 50, MORE HIGHLY MAGNIFIED. (After 



Hubrecht.) 



emb. ect.y embryonic ectoderm ; pp, primitive 

 entodermic plate ; y.s., yolk-sac ; mes, ventral 

 mesoderm. 



Sections of the primitive streak show 



