68 BULLETIN OF THE 
In freshly laid eggs this layer is in contact with the vitelline mem- 
brane, but early becomes separated from it by the contraction of the 
vitellus. The perivitelline fluid which makes its appearance during this 
process is coagulable by heat and is also stainable. At first the contrac- 
tion of the egg is uniform on all sides, but soon it takes place more 
rapidly on one side, thus giving rise to a flattened surface (compare 
Fig. 5), upon which the ventral plate is afterwards established. Through 
the pressure of this contraction the blastema is moulded upon the periph- 
eral yolk corpuscles into regions that correspond in position and size 
with the underlying corpuscles. Owing to mutual pressure these regions 
become regular hexagonal areas, (Pl. V. figs. 24, 25,) and resemble the 
subsequently formed cells of the blastoderm. The absence of nuclei is 
the fundamental feature that at once distinguishes them from the blasto- 
dermic cells, though they have frequently been mistaken for such on the 
supposition that the nuclei were obscured. 
The division of the blastema into areas as described above is a very 
early phenomenon. At the time of my first observations a number of 
faintly marked areas had already made their appearance at the more 
active (animal) pole. At this time they could not be detected upon the 
opposite hemisphere; but after a short interval they also made their 
appearance there in isolated patches ; finally they covered the entire sur- 
face of the egg. At the outset the boundary lines of the areas are very 
faint, but they become more distinct as the contraction of the vitellus 
continues. In some places the yolk corpuscles become separated from 
the blastema by a more rapid contraction of the interior protoplasm, and 
then the polygonal areas in such regions remain only partially outlined 
and incomplete, as described and figured by Balbiani (73, Fig. 2). 
After the areas are definitely formed the yolk corpuscles sometimes 
shift their original positions, and thus cease to coincide with the areas, 
since the latter do not at the same time undergo corresponding changes. 
The next alteration in the surface makes its appearance only after the 
lapse of a considerable interval (twelve to forty-eight hours) ; this led 
Balbiani to assert, erroneously, that the egg is undergoing a period of 
rest. Sections show on the contrary, that the interim is one of great 
internal activity, during which repeated divisions of the nuclear sub- 
stance lead to the formation of numerous cells which migrate towards the 
periphery. The appearance of some of these at the surface marks 
the beginning of new surface changes. The cells thus emerging from 
the yolk constitute the primary blastoderm ; they first appear in the 
interspaces between the yolk corpuscles, but often migrate afterwards to 
