HUMAN AND MAMMALIAN 27 



througli the suhstantia gelatinosa, and this is perhaps true in the lowest sections 

 (Fig. 3"), in which it is, however, probable that the roots passing in front of the 

 caput cornu are the uppermost of the spinal accessory. In sections higher up, all 

 the roots penetrate the caput, a circumstance which Stilling was the first to point 

 out as proving the resemblance between the vagus and posterior spinal roots. 



The roots of the vagus enter the anterior spur of the nucleus, just in front of the 

 great bundles of longitudinal fasciculi, which separate it from the posterior process, 

 their fibres being distributed much on the same plan as those of the accessory. 

 Their course can be studied in Plates X, XI, Figs. 38 and 40. A portion of the 

 fibres belonging to the great bundle of roots ( F), enter deeply into the nucleus, 

 becoming sooner or later united with cells which are very numerous and much 

 crowded together (Plates XIII, XV, 3«, 4% 6% 21% 22"). Another portion of the 

 roots, as wiU be seen in Fig. 40, turn backwards and pass out among the longitu- 

 dinal bundles {H), or form loops around them, while some pass stiU further back 

 into the posterior spur. Of the bundles which pass forwards, without entering the 

 ceUs of their own nucleus, Clarke makes the following statement : " A separate 

 bundle turns inwards, and after sending forwards in succession a number of single 

 returning fibres, which wander through the network of the lateral column, proceeds 

 through the side of the hypoglossal nucleus, where its fibres mingle with those of 

 the hj'poglossal nerve — I may almost atfirm that some at least are continuous witli 

 the cells. Such is the course I have repeatedly observed in man ; and in the sheep 

 and ox I can show, without any difficulty, that while some of the fibres of the last 

 mentioned inner bundle are apparently continuous with those of the hypoglossal 

 nerve, others pass inwards to decussate at the raphe."^ 



On the other hand, Schroder van der Kolk^ has endeavored to show that there is 

 no dii-ect decussation either of the sensitive or motor nerve roots of the medulla, 

 but that the decussation of both is produced through the intervention of cells. 



This statement is, I think, incorrect in regard to the ,spinal accessory and vagus 

 roots, the course of which I have had repeated opportunity of studying in very many 

 specimens, both from the human medulla and that of various mammals. ' 



The course of the spinal accessory roots has been described and figured with 

 very great accuracy by Clarke,^ with whose observations my o\vn have agreed very 

 fully,'* proving to my entire satisfaction, that while a part of the roots enter the 

 nucleus and proceed directly towards the cells, quite a large bundle turns forwards, 

 and passing in front of the hypoglossal nucleus, for the most part without entering 

 ceUs, decussates at the raphe with a similar bundle from the opposite side. 



The same is true of the vagus roots, but here the bundles turning forward to 

 decussate are much larger and more conspicuous than is the case with those proceed- 



1 Philos. Transactions, 1858, 263—4. 



" Medulla Spinalis and Oblongata, 96, 185. 



' Philos. Transactions, 1858, pi. xviii, fig. 35. 



* A sketch drawn from one of my own specimens was so exact a reproduction of Clarke's figure in 

 all important particulars as to render the publication of it, as well as of my own observations on the 

 course of the spinal accessory roots, quite superfluous. 



