If this latter hypothesis be accepted, then it is necessary to explain why parts of that tube, such as the roof of the fourth ventricle, the choroid plexuses of the various ventricles, which are parts of the original roof inserted into the ventricles, are not composed of nervous material, but form simple single-layered epithelial sheets, which by no possibility can be included among functional nervous structures. The upholders of this hypothesis can only explain the nature of these thin epithelial parts of the nervous tube in one of two ways; either the tube was originally formed of nervous {39}material throughout, and for some reason parts of it have lost their nervous function and thinned down; or else these thin epithelial parts are on their way to become nervous material, are still in an embryonic condition, and are of the nature of epiblast-epithelium, from which the central nervous system originally arose. The first explanation is said to be supported by embryology, for at first the nerve-tube is formed in a uniform manner, and then later, parts of the roof appear to thin out and so form the thin epithelial parts. If this were the right explanation, then it ought to be found that in the lowest vertebrates there is greater evidence of a uniformly nervous tube than in the higher members of the group: while conversely, if, on the contrary, as we descend the vertebrate phylum, it is found that more and more of the tube presents the appearance of a single layer of epithelium, and the nervous material is limited more and more to certain parts of that tube, then the evidence is strong that the tubular character of the central nervous system is not due to an original nervous tube, but to a non-nervous epithelial tube with which the original nervous system has become closely connected. The comparison of the brain region of the different groups of vertebrates (Fig. 19) is most instructive, for it demonstrates in the most conclusive manner how the roof of the nervous tube in that region loses more and more its nervous character, and takes on the appearance of a simple epithelial tube, as we descend lower and lower; until at last, in the brain of Ammocoetes, as represented in the figures, the whole of the brain-roof, from the region of the pineal eye to the commencement of the spinal cord, is composed of fold upon fold of a thin epithelial membrane forming an epithelial bag, which is constricted in only one place, where the fourth cranial nerve crosses over it.