in accepting a generalized interpretation, the more so since we have Henle’s observation of the similar entrance of a _branched_ process into the root.[149] Now it must be remembered that the branched process is by no anatomist at present regarded as the origin of the axis cylinder; so that if it can enter the root without being the origin of a nerve-fibre, we are not entitled to assume that the entrance of the unbranched process has any other significance (on this head compare § 145), especially when we reflect that no trustworthy observer now professes to have followed a nerve-fibre of the posterior root right into a multipolar cell. Figures, indeed, have been published which show this, and much else; but such figures are diagrams, not copies of what is seen. They belong to Imaginary Anatomy.[150] The relation of the cell-process to the nerve-fibre will be discussed anon. [Illustration: Fig. 19.--_Anastomosing nerve-cells_ (after Gratiolet). _a_, body of the cell; _c_, process of uniting two cells; _d_, branching process. ] 125. A word in passing on the contradictory assertions respecting the anastomosis of nerve-cells. That the gray substance forms a _continuum_ of some kind is certain from the continuity of propagation of a stimulus. But it is by no means certain that one cell is directly united to its neighbor by a cell-process. Eminent authorities assert that such direct union never takes place; others, that it is a rare and insignificant fact; others, that it is constant, and “demanded by physiological postulates.” I will not, in the presence of distinct affirmations, venture to deny that such appearances as are presented in Fig. 19 may occasionally be observed; the more so as I have myself seen perhaps half a dozen somewhat similar cases; but it is the opinion of Deiters and Kölliker that all such appearances are illusory.[151] Granting that such connections occur, we cannot grant this to be the normal mode; especially now the more probable supposition is that the connection is normally established by means of the delicate ramifications of the branching processes. Imaginary Anatomy has not been content with the cells of the anterior horn being thus united together, to admit of united action, but has gone further, and supposed that the cells of the posterior horn, besides being thus united, send off processes which unite them with the cells of the anterior horn--and thus a pathway is formed for the transmission of a sensory impression, and its conversion into a motor impulse. What will the reader say when informed that not only has no eye ever beheld such a pathway, but that the first step--the direct union of the sensory nerve-fibre with a cell in the _posterior_ horn--is confessedly not visible?