No. 5-] HABENULAR TRACTS OF ACIPENSER. 227 



are much more complex than in mammals. They may receive 

 from two or three to a large number of olfactory fibers, and 

 the size of the glomerulus is probably determined by the num- 

 ber of olfactory fibers entering it. Each glomerulus may be 

 supplied by a single large or small bundle of fibers, or by sev- 

 eral single fibers or small bundles of fibers coming, it may be, 

 from widely separated bundles of the olfactory nerve. Each 

 olfactory fiber usually has from two to five end-twigs produced 

 by dichotomous divisions, but frequently the branching is 

 irregular, and it may be very complex. I have found no indi- 

 cation of anastomosis or the formation of a network in these 

 endings. To the central portion of the glomerulus dendrites 

 may be contributed by any one or several of the eight forms 

 of cells described above. The largest part of a typical glomer- 

 ulus is formed by the end-branching of the mitral cell dendrites, 

 but almost all the glomeruli in Acipenser are formed in part 

 by the dendrites of other cells. The whole makes up a very 

 complicated mass of interwoven nerve twigs among which 

 ramify the processes of glia cells. The essential condition for 

 the transference of nerve impulses in the glomerulus is the 

 contact of the tips of the olfactory fiber end-branches with the 

 dendrites. There is in Acipenser no evidence of continuity of 

 nerve substance between olfactory fibers and dendrites. Be- 

 sides these typical glomeruli, there are many small ones whose 

 central portion is formed wholly by the dendrites of cells other 

 than the mitral cells. 



I have thus recognized eight forms of nerve cells in the 

 olfactory lobe. On the whole, these cell forms are distinct and 

 present well-marked characters. However, there are to be 

 found occasional intermediate forms: between the cells a and 

 c, b and c, b and e, c and/ (Fig. i). Only the large and small 

 mitral cells (and the granule cells ?) can with certainty be com- 

 pared with cell forms heretofore described in other vertebrates. 

 The olfactory lobe of other fishes (especially Cyclostomes and 

 Elasmobranchs) and of Amphibia needs to be investigated by 

 the Golgi method and the lobe of higher vertebrates reexamined 

 with reference to the occurrence of the other cell forms. The 

 only work done by modern methods on the olfactory lobe of 



