framework in between the cells of the nervous layer. It is quite possible that we ought to regard the nervous layer of epiblast as comparable to the germinal cells of His rather than to the fully-developed neuroblasts.

If this is correct, we then must conclude that in the frog there is a very early separation of the neuroblastic from the spongioblastic elements.

Comparison between Rabbit and Frog.

Is it then possible that the condition in the rabbit has its parallel here in the frog? It is true that the epiblast is double only over a certain area of the embryo in the rabbit, whereas in the frog it is double throughout.

In the frog the nervous layer soon becomes much thickened along the future dorsal surface of the embryo, and over the rest of the embryo the nervous layer becomes reduced to a layer of one cell only in thickness, like the epidermic layer.

Now, although it is extremely difficult to trace the history exactly, I am almost sure that the area over which the inner layer of epiblast cells in the rabbit is found, corresponds to that area in the frog over which the nervous epiblast remains thick, or becomes thicker-i. e. to the neural plate.

In the frog the whole of the neural plate does not become folded up to form the neural tube, but the outer lateral portions of the anterior part remain outside of the tube, giving rise to the ganglia of the anterior cranial nerves.

I have endeavoured elsewhere to bring evidence to show that the epiblastic wall of the anterior part of the body of the rabbit embryo includes more than the double-layered part of the embryo, i. e. more than the so-called "embryonic disc."

Whether the" embryonic disc" goes to form the neural tube and ganglia of anterior cranial nerves as in the frog, or whether it forms only the neural tube, I have no evidence to offer.

The embryonic disc is precisely in the same position relative to the primitive streak as is the neural plate to the blastopore of the frog.

I am well aware that the epiblast is not at first double in all

mammals. For instance in the opossum, according to Selenka, it is a single layer. But it also is not always double in the Amphibia. In Triton it is at one time only a single layer of one cell in thickness. Why there should be this early differentiation into spongioblastic and neuroblastic elements in one and not in another so comparatively closely allied animals it is not easy to guess. Possibly it may be that, since the spongioblastic elements of the nervous system are the first to show activity of growth in the nervous system (His, and above), then in those animals in which, owing to various individual causes, the epiblast is many-layered, the outermost layer of "cells" or centres of activity being, by reason of its external position more favorable to processes of respiration and so in a condition more favorable to active growth, it will be this layer of epiblast that will take on itself the earliest phase in the further development of the nervous system.

Although I think the rabbit's condition can be quite well explained without reference to the above, on the other hand there may be some deeper meaning in it, for which reason I have thought it best to make notice of the condition in the Anura.

To make this parallelism complete and certain it is necessary to show that in the rabbit the cells derived from the outer layer give rise to the spongioblastic tissue, and those derived from the large inner layer cells to the neuroblastic tissue. This I cannot do. After a while the cells which have originated in each of the layers become equally active, but I have as yet been unable to trace their fate respectively.

DESCRIPTION OF PLATE 18,

Illustrating Mr. Richard Assheton's paper, "On the Phenomenon of the Fusion of the Epiblastic Layers in the Rabbit and in the Frog."

LIST OF REFERENCE LETTERS.

PI. Stalk of pineal gland.

E. EP. Epidermic
PR. Processes of

AU. Auditory vesicle. E. NE. Nervous epiblast. epiblast. NE. Neuroblast. epidermic layer cell. T. E. Band of nerve-fibres passing from ganglion habenula to the ventral side of the brain.

FIG. 1.-Transverse section through a portion of the neural plate of a frog embryo (Rana temporaria) during the process of folding up of the neural plate. Unstained. x 165.

FIG. 2.-Transverse section through corresponding region, but at a slightly later period, of a frog's embryo. Unstained. × 165.

FIG. 3.-Transverse section through the corresponding region after the separation of the neural tube from the skin. Tadpole, 34 mm. Unstained. X 165.

FIG. 4.-Transverse section through the spinal cord of the same specimen as Fig. 3. Unstained. × 165.

FIG. 5.-Transverse section through ganglion habenula of a frog tadpole of 13 mm., showing deeply pigmented neuroblastic processes. Aniline blueblack. x 350.