CELLS OF THE NERVUS TERMINALIS OF THE RAT 107 



so-called oKactory ganglion. The two regions are shown in 

 figures 4 and 5 — photographs of the same embryo, age 15f days, 

 V. Rath technique, and under the same magnification (X 780). 

 The cells of the olfactory region — those clearly neuroblastic in 

 nature — are in general rounded, with rounded or slightly oval 

 nuclei; their cytoplasm is relatively abundant. Those in the 

 'olfactory ganglion' are situated mainly on its medial aspect and 

 are traceable caudad along the bulb, along what are clearly 

 nervus terminalis strands. The cells of the sphenopalatine 

 ganglion still tend to lie in strand-like aggregates ; they are more 

 elongated with nuclei generally markedly oval and with rather 

 scanty cytoplasm. The two varieties are readily distinguishable. 

 In brief, the process has been followed up to and including 

 seventeen-day embryos, and no evidence of an origin, other than 

 that earliest observed from the olfactory epithelium and con- 

 tinuing into fairly late embryonic life from the same source, 

 has been obtained for the ganglion cells of the nervus terminalis. 

 I have likewise given some attention to the suggestion of Larsell 

 ('19) of an origin from the neural tube, and must confess similar 

 negative results. The growth appears to me to take quite the 

 opposite direction, no contact of ganglion cells with the brain 

 wall occurring until late stages, when cells are carried backward 

 with the growing nerve rootlets. To be sure, this study does 

 not carry the problem into postembryonic or even into the 

 really late stages of fetal life, but evidence obtained from cell 

 counts indicates a decrease rather than an increase in those 

 cells regarded as neuroblasts, this being in essential agreement 

 with the observations of Huber, and doubtless likewise subject to 

 the same sources of error. In view of this apparent decrease it 

 would be rather difficult to assume late additions to the gan- 

 glionic aggregates of the nervus terminalis. The writer sees 

 therefore no reason for reversing his earlier decision or the 

 decision reached by certain previous investigators and must 

 consider the derivation of ganglion cells of the nervus terminalis 

 from sources other than the epithelium of the olfactory sac, i.e., 

 from those sources suggested as possible earlier in this paper, as 

 resting upon evidence as yet purely gratuitous. 



THE JOURNAL OF COMPARATIVE NEUROLOGT, VOL. 32, NO. 1 



