378 Julia B. Platt 
nothing other than the sum of the embryonic ceils which have been 
left unused in the formation of epithelial organs (loc. cit.). We have 
here ground for a seemingly sharp distinction between structures 
formed from epithelia which may be traced directly to one or other 
of the primitive germ-layers, and structures of mesenchymal origin, 
in the formation of which any or all of the germ-layers may 
participate. This distinction, however, is not as definite as it at 
first might seem. Cells of the mesenchyma, primarily derived from 
epithelial layers, may again secondarily associate themselves in the 
formation of epithelia. Thus, the wall of the »anterior« head-cavity 
in the Selachii is formed from scattered cells, i. e. mesenchyma, of 
endodermie origin, yet, at the stage of its highest development, the 
wall of this head-cavity is perfectly epithelial in structure, differing 
apparently in no wise from the wall of the neighboring mandibular 
cavity, which traces its origin directly to the primitive mesodermic 
germ-layer, and ultimately gives rise to the mandibular muscles. 
Moreover, the homologue of an organ derived in one Vertebrate 
from epithelial tissues, may, in another Vertebrate, be derived from 
the mesenchyma. Thus, the eye-muscles which develop in the epi- 
thelial walls of the praemandibular, mandibular, and third head- 
cavities in the Selachii, are formed in the Amphibia in the 
mesenchymal tissue median to the optic vesicles. These instances 
serve to show that the distinction between epithelial and mesenchymal 
tissues is indefinite, and that the division of the mesoderm on the 
basis of this distinction lacks wide morphological significance, 
although most conveniently descriptive of actual histological condi- 
tions. Doubt which attaches to the origin of one part of the middle- 
layer, does not leave the other part altogether unaffected, but rather 
brings into question the morphological value of the whole. 
Is there.in the Vertebrata a median germ-layer, whose cells 
have a common origin, and which gives rise to definite systems of 
organs? v. KÖLLIKER (’84, page 201) answers this question by ably 
defending the view that it is no longer possible to associate each 
germ-layer exclusively with one or more tissues, but that each layer 
is capable of giving rise to all of the principal tissues. 
Observed processes of growth, however, seem to warrant the 
assumption that, in the Vertebrata, at the time of gastrulation, 
the cells of the inner layer have been so far differentiated that they 
are no longer capable of producing nervous tissues, and that the cells 
of the outer layer are no longer able to fulfil the digestive functions 
