640 MANUAL OF HISTOLOGY. 



perpendicularly by the fibres of Muller (when seen in vertical section) 

 is found to be composed mainly of a dense connective-tissue network. 

 It contains, as was found by Muller and Schultze in fishes, and as Koel- 

 liker recognised in the mammalia, a superficially expanded network (with 

 nuclei in the nodal points), formed by the union of flattened stellate 

 cells. This exists, according to Krause, as a single layer in all verte- 

 brates likewise, and has been named by him membrana fenestrata, and 

 declared to be an extremely important " boundary structure " for the 

 retina. 



At present but little is known as to the arrangement of the nervous 

 fibre elements in the intergranular layer. According to Schultze, there 

 exists here an interlacement of the finest fibrill?e running horizontally and 

 obliquely (fig. 596, B, d), formed by the splitting up of the rod and cone 

 fibres already mentioned ( 316). In the opinion of Hasse, however, the 

 two side processes only of the latter run a short distance in an oblique 

 direction, the other middle fibre sinking perpendicularly into the stratum 

 intergranulosum. 



(5.) The stratum granulosum internum or internal granular layer is 

 usually (0'03-0'04 mm.) of less depth in man than the external. Its 

 " granules " are also somewhat larger, and more distinct in outline. 



From the observations of Vintschgau, Muller, and Schultze, it would 

 appear that two kinds of elements may be distinguished here. In the 

 first place (fig. 596, B, /), we meet with roundish cells remarkable for 

 their distinct and brilliant outline, large nuclei with nucleoli, and small 

 amount of cell-body. They are usually spoken of as bipolar, i.e., giving 

 off a process at either end. These processes are extremely delicate, but 

 that directed outwards towards the intergranular layer is of considerably 

 greater diameter than that springing from the opposite pole (Schultze) 

 in the second place, we find pale-edged oval nuclei with large nucleoli (A,e f ). 



The latter elements belong to the connective-tissue of the part. They 

 are not, however, as was formerly supposed, imbedded in the fibres of 

 Muller, but only firmly seated upon the latter, enveloped in masses of 

 delicate spongy tissue, and may be regarded as the central points of cell- 

 like bodies. In number they fall far short of the first species of granule. 

 For the rest, the connective-tissue framework of the internal granular 

 stratum presents the same constitution as that of the external. 



Here also we are ignorant of the course of the nerve fibres. As in the 

 last layer treated of, so here strong fibres, corresponding to the axis 

 cylinders of the cones, are wanting in the stratum granulosum externum. 

 We can only expect to find here primitive fibrillae, or plexuses of the 

 same. The fine threads which spring from both poles of the " granules," 

 giving to the latter the appearance of small bipolar ganglion-cells, are 

 probably fragments of such. After a short course, however, they are 

 lost to sight. In some instances such a thread, springing from a cell 

 situated high up, extends upwards into the inter-granular layer. Again, 

 springing from cells seated low down near the molecular layer, filiform 

 processes may be seen which sink into the latter. 



(6.) The stratum moleculare, or finely granular layer (Bg), presents very 

 much the appearance of the delicate molecular matter already encountered 

 in the grey substance of the brain and spinal cord. Under high micro- 

 scopic power, it is found to be made up of fine spongy tissue. As has 

 been before mentioned ( 315), this layer (in man 0'03-0'04 mm. thick) 

 is traversed vertically by the fibres of Muller. It appears to contain a 



