while at other times, the oily contents of the latter reach quite up to the corpuscle, and seem either to become blended with the granular contents of the same, or are pro longed into the clear space. Wagner* be lieves that the normal relation betw een the nerve-tubes and corpuscles is, that each pri mitive nerve-fibre coming from the centre, retains its appearance of double contour up to the ganglion-cell, where its contents are interrupted by the finely granular matter of the latter; at the peripherical extremity of the cell, the nerve-fibre commences in a quite similar manner. Thus then, according to Wagner, the oily contents of the nerve-tube cease on reaching the ganglionic corpuscle. Biddert, on the other hand, regards the clear white space between the cell-wall and the granular contents as a thin sheet of nervous matter, which serves as a connecting medium between the contents of the nerve-tube on either side of the ganglionic corpuscle.
As regards the srnpathetic ganglia in this animal, they appear in some parts to be ahnost entirely composed of a fibrous structure and of a quantity of granular matter resembling, as Wagner and Robin observe, the gray granular matter of the nervous centres, and containing a number of bodies of a brownish yellow granular appearance, which do not dis It has been already seen that it is almost certain that all the ganglionic corpuscles oc curring in the ganglia on the posterior roots of the spinal nerves, at least in this animal, belong to the bipolar variety ; it remains to inquire whether all the nerve-fibres in the posterior root are connected with ganglionic corpuscles. When the ganglion, after addition of dilute solution of soda, is examined with a power of about 30 or 40 diameters, itis observed that while many of the fibres soon disappear among the ganglionic corpuscles, several of them can be traced from the point at which they enter the ganglion almost to its opposite extremity, without being connected with cor puscles ; but I have never been able to trace them in this manner quite past the ganglion. Wagner, who counted the nerve-tubes con tained in the posterior root of the nerve and also the ganglionic corpuscles, found that the number of each corresponded pretty closely, so that he believes each of them is connected with one of these bodies.
Invertebrata.— The ganglionic corpuscles in the ganglia of the invertebrata appear to be the same in their essential characters as those of the vertebrate animals. Will* re cognises two kinds of ganglion-corpuscles in the lower animals. The one he describes as consisting of a membrane and nucleus, the space between the two being occupied by a clear transparent fluid, which becomes granular on the addition of water ; in the other variety there are imbedded in the clear transparent fluid numerous small round cells in which no nucleus is visible. The cells belonging to the former variety have always but one process attached to them which consists of a single tube, presenting no division so far as it can be traced, and thus corresponding to the uni polar variety of corpuscle. In the second kind of corpuscles there are several such pro cesses present ; the processes attached to some of these cells all run in one direction ; in others they pass off at either extremity and run in opposite directions. In the leech, according to Brucht, there are also two kinds of ganglion corpuscles. The one variety are round and are apolar ; the others are con nected with nerve-fibres. The latter are situ
ated towards the lower part of the ganglia, and are more numerous than the former: they are more or less pyriform, their wider ex tremity being directed outwards ; their nar rower, terminating in a process, is directed towards the ganglia and the nervous cord. Peripherical ganglia, consisting of from one to six or seven cells, are always found at the points where the branches of the nerves di vide. Ganglionic corpuscles were also seen by him in the interior of the nerve-tubes, and corresponding to the view taken by Biddert of the constitution of the bipolar ganglionic corpuscle. Apolar and unipolar cells have also been described by Hannover and Leydig in several other invertebrate animals.
From the fact that in such animals as the torpedo and ray, where the ganglionic cor puscles are easily isolated from each other, they are all found to belong to the bipolar variety, Wagner, Robin, and Bidder believe that all the ganglionic corpuscles in other animals are also bipolar. Kolliker, on the other hand, while he admits that the bipolar cell is most frequent in the fish, maintains that the opposite is the case as regards the higher animals, most of the corpuscles in them belonging either to the apolar or uni polar varieties; and so far as actual observa tion goes, the views of Kolliker seem to be perfectly correct, inasmuch as, while apolar and unipolar cells are very frequently seen in these animals, the bipolar variety has been seen very seldom. It is possible, however, that many of these unipolar and apolar cells may, as Wagner and Bidder, &c. hold, be really bipolar cells, one or both nerve-tubes having been broken off during the manipula tion required for submitting them to examina tion. In the spinal ganglia of the ray the cells are very easily isolated from each other, whereas in the abdominal ganglia it is very difficult, owing to the amount of surround ing fibrous structure, to isolate them. Now in the former only bipolar cells are seen, whereas in the latter, most of the cells, when isolated, appear to be unipolar and apolar, although it would appear from the observa tions of Wagner and others, that they are all bipolar, like those in the spinal ganglia. In the higher animals, especially in the mam malia, the ganglionic corpuscles are isolated from one another with as much difficulty as those in the abdominal ganglia of the skate ; and hence the probability that many at least of the unipolar and apolar cells which are seen in them, belong to the bipolar variety in reality. On the other hand that apolar and unipolar ganglion-corpuscles really exist, and that too in considerable numbers, in the gan glia of the higher animals, and also in those of the invertebrata, seems to be shown by numerous observations on the smaller gan glia, where no preparation is required, and where, consequently, the above source of fallacy cannot intervene. In the sympathetic cord of the frog, according to Valentin*, groups of ganglionic vesicles may be observed, without a single nerve-fibre connected with them : Ludwig t has also observed in the au ricle of the frog,'s heart small ganglia in which there were eleven ganglionic corpuscles, and only four or five nerve-tubes ; in a nerve passing to the bladder of the frog, and con sisting of only two nerve-fibres, Valentin counted as many as seven ganglionic cor puscles, while another, consisting also of only one or two nerve-fibres, was surrounded by twenty-four ganglionic corpuscles.