No. I.] THE VERTEBRATE EAR. ^21 



question must be answered in the affirmative, and that each 

 fibril acts independently, and as a unit ; in other words, as an 

 auditory hair, and not as a fibril of a hair, or one of a bundle 

 of hairs. For I believe, though at present there is not suffi- 

 cient morphological basis to give the view much weight, that 

 each capillary fibril of the auditory cell of the human cochlea 

 has independent connection with the brain through its own 

 nucleo-neural filament and its own fibril of the transmittina: 

 nerve. If this were so, it would be possible for 770,000 sepa- 

 rate impulses to reach the brain from the cochlear organs as 

 the result of a single stimulus exciting the external air. 



Loewenberg's net is a structure which was originally de- 

 scribed by the investigator whose name it bears. It is appar- 

 ently a network of fibres covering the membrana tectoria 

 wholly or in part. It was found to be especially well developed 

 on the outer surface of the tectorial membrane. For other his- 

 tological details consult the historical account of the membrana 

 tectoria in a preceding paragraph. This net, the membrana 

 tectoria, with which it is intimately connected, and the mem- 

 brana reticularis may be justly said to have constituted the 

 three mysteries of cochlear anatomy, if one may be allowed to 

 overlook those cochlear phantoms, the spiral, i.e. longitudinal, 

 bundles of nerve fibrillas of Deiters, Nuel, Retzius, and others. 



Since I have frequently found this net, and since its nature 

 was for a long time uncertain, I shall describe its ordinary ap- 

 pearance briefly as a preliminary to the account of its true 

 nature, as deduced from a large series of nets. The net usually 

 lies closely adherent to the surface of the membrana tectoria, 

 and presents the appearance shown in PI. II, Fig. 5. The 

 coarser meshes are always nearest the inner edge of the hair 

 band, while the coarser fibres are in the outer edge. These 

 fibres or threads are very much like the fibres of the hair band 

 in appearance and give the same physical and chemical reactions 

 with few exceptions ; e.g. the fibres of Loewenberg's net not 

 unfrequently stain more deeply than the hairs of the hair band. 

 Although the net is usually closely applied to the surface of the 

 hair band in microscopical preparations, one often finds it more 

 or less raised from this surface in fresh preparations, or it may 

 be entirely absent from the organ. I have preparations of rat, 

 rabbit, dog, pig, and opossum (PI. XI, Fig. 2) which do not 



