In accordance with the view adopted by Wagner, that all the ganglionic corpuscles are bipolar, the nerve-tube connected with either extremity of the cell running in opposite directions, one towards the centre, the other peripherically, Robin believes that all nerve tubes arise exclusively from the brain and spinal cord ; neither the spinal ganglia nor those of the sympathetic give origin to nerve tubes; the ganglion-cells are merely organs developed upon the nerve-tubes, between their central and peripherical termination, and several such may be present on a single nerve fibre during its course. From what has been already stated, however, it seems probable that unipolar as well as bipolar cells exist in the ganglia, and consequently that nerve tubes do originate in them. That nerve-fibres arise in the ganglia, is further shown by the accurate measurements of Volkrnann and Bidder of the nerves passing to and those leaving the ganglia. The ciliary, Gasserian, and spinal ganglia in the frog were found by them to give off a far greater number of fine nerve fibres on the one side than they received on the opposite side. So also in the septum between the auricles of the frog's heart Bidder has seen small ganglia, which gave off' on the one side eight nerve-fibres more than they received on the other side. The obser vations of Bidder and Volkmann have been confirmed also by KoHiker. Engel*, moreover, describes a peripheral ganglion, to which no nerve-fibres passed, while a number of fibres left it; an observation which, if correct, places beyond a doubt the question as to the origin of nerve-fibres in the ganglia. The ganglion in question he describes as being pear-shaped, and about 0.096 of a line in diameter ; it oc curred in the perichondrium of the tracheal cartilage, and consisted of fourteen ganglionic corpuscles, with seven efferent nerve-fibres, each measuring about 0.0012 of a line in diameter. Even in regard to the bipolar gan glionic corpuscles, it does not appear to be at all certain that they are all merely organs deve loped on the course of a nerve-fibre arising in the brain and spinal cord. On the contrary, it would appear that several of the cells be longing to this variety must also be regarded as giving origin to nerve-fibres in the same manner as the unipolar cell. Thus Bidder has seen bipolar cells, the nerve-tubes con nected with which did not run in opposite directions, one towards the brain and spinal cord, the other towards the periphery, but both ran in the latter direction (fig. 291.) : so also Stannius, as mentioned by Kolliker, has seen in the ciliary ganglion of Trigla, a bi polar cell, both nerve-fibres of which were directed peripherically. The same observer has also seen ganglionic corpuscles in the fish which gave origin to or had three nerve-tubes connected with them.
That most of the bipolar cells are, how ever, as Robin maintains, organs developed on nerve-fibres of cerebro-spinal origin, in their course towards the periphery, there is no reason to doubt; and moreover that several of these may occur in the course of a single fibre between its central and peripherical ter mination is also shown by the observations of Stannius on :the fish, and by Valentin on the frog. Wagner has also observed two ganglion-corpuscles occurring in the course of a single nerve-fibre, at short distances from one another.
Robin* divides the ganglionic corpuscles into tviro distinct classes, a larger and a smaller : the larger he finds always occur on broad nerve fibres, or fibres of animal life, while the smaller are always connected with nerve-fibres be longing to the finer variety, or fibres of or ganic life ; and in this way, according to him, we have a good mark by which to distinguish the animal from the organic nerve-fibres. In
the ray, according to Robin, the larger variety of corpuscles measure 0.095 to 0.150 nim. in diameter, are spherical, and often flat at both poles ; the smaller measures 0.080 to 0.115 mm. in length, and 0.050 to 0.070 mm. in breadth, and are commonly oval. .In the larger cells there is a layer of clear round bodies, without nuclei; in the smaller gangli onic corpuscles the outer membrane is finer, and each of the cells, on their inner sur face, , is provided with a central dark nu cleus. Bidder -I- also agrees with Robin in separating the ganglionic corpuscles into two groups. In the pike the one set measure 0.042'", while the other set do not measure more than 0.018'": the former chiefly occur in the ganglia of the cerebro-spinal nerves, the latter in the ganglia of the sympathetic ; the former are always connected with broad fibres, the latter with fibres belonging to the fine variety. The views of Robin and Bidder are opposed by Kiilliker, Valentin, and ap parently also by Wagner. The latter admits that in general the ganglionic corpuscles are smaller than those occurring in the spinal ganglia, and that the smaller corpuscles have, as Robin observes, an oval shape, while the larger are more or less spherical : there are, however, according to him, cases where broad nerve-fibres are seen passing off' from small cells, and where the large cells are connected with small or narrow fibres. Sometimes, in deed, the ganglionic corpuscle has a narrow tube on one side, and a broad one on the op posite side (seefig. 290.); and sometimes the broad, sometimes the narrow, runs peripheri cally. Stannius has, as mentioned by Kiilliker, observed in Petromyzon cells present, of the fibres connected with which the one was six times broader than the other. Although, how ever, there does not appear to be a distinct de marcation between the ganglionic corpuscles belonging to the two sizes, there can be little doubt that the cells occurring in the sympa thetic ganglia are generally smaller than those occurring on the cerebro-spinal nerves, both in the fish and also in the higher animals. The larger cells in the spinal ganglia of the ray appear, as Robin states, to be (in general at least) connected with broad fibres, while the smaller cells are connected with narrow fibres : this, however, does not appear to be invariably the case. In the sympathetic gan glia there are sometimes seen connected with narrow fibres cells as large as some of those in the spinal ganglia, which are connected with broad fibres. Moreover, as already stated, there appear to be transitional sizes between the larger and smaller variety of cor puscles.Hiker also calls attention to the fact that small ganglionic corpuscles occur in other parts than in the sympathetic, as for example those in the brain and spinal cord. It would seem, then, that just as the finer variety of tubular nerve-fibres cannot be re garded as characteristic of the sympathetic system, so also the smaller variety of gan glionic corpuscles cannot be reg,arded as pe culiar to it either.