Herrick-Coghill, Nerve Endings in the Skin. 5 1 



ficulties in the way of a correct interpretation of the appearances 

 presented by sections, 



A discussion of the theoretical bearings of these facts and 

 further details must be deferred to the second part of this paper. 



Since writing the above we have been able to settle several 

 points previously in doubt. None of our preparations of the 

 skin of amphibians gave unambiguous results for the glands of 

 the skin. We have at last succeeded in securing excellent in- 

 tm vitam impregnations in the toad ( Bufo sp.) in which it is 

 easy to trace the non-medulated fibers from the plexus ectad of 

 the corium, and also from that entad of it, into the most inti- 

 mate connection with the superficial walls of the glands, which 

 in this species are very large and highly functional. The fibers 

 are of small caliber but are excessively numerous and envelop 

 the whole gland in what at first looks like a closely woven re- 

 ticulum, but a close study shows that the appearance of a retic- 

 ulum is due to the repeated dichotomous branching of a large 

 number of distinct nerve fibers. These fibers cross at slightly 

 different levels and there is no doubt in most cases of the com- 

 plete distinctness of the fibers as they cross. Upon these 

 fibers are frequent varicosities which may be due to imperfec- 

 tions of the process or may be the points of attachment of the 

 fibers upon the cells of the gland. Of course this method does 

 not admit of determining the exact relation of the nodosities to 

 the several cells, but there can be no doubt of the existence of 

 a very intimate and necessary connection. One is forcibly 

 struck by the close resemblance of this periglandular felting to 

 the perigemmular reticulum described by many authors in the 

 case of the sense buds. The latter is, as we have before in- 

 sisted, entirely distinct from and totally unlike the intragem- 

 mular endings in distinct cells which may be demonstiated by 

 a wide range of independent methods. 



The same preparations used in the earlier parts of this paper 

 have also afforded to a more extended study a number of sat- 

 isfactory views of the connection of the ganglion cells of the 



