1 68 Journal of Comparative Neurology. 



throughout the extent of the medulla spinalis, the cell-bodies, 

 within certain limits, vary in size for the different localities — 

 those of the thoracic region having an average diameter less 

 than those of either the cervical or lumbar. Kaiser's results must 

 differ from the author's chiefly from the fact that looking for ex- 

 tremes he employed a much larger number of cell bodies from 

 which to choose his largest. 



Waldevek ('89) measured a small number of cell bodies 

 of the Ilird and IVth cervical segments of a two year old male 

 gorilla, and from his records an average mean diameter of only 

 32 micra is obtained. His measurements deal with the ventral 

 group of the columna anterior and, unless the gorilla differs 

 from the mammals with which the author is acquainted, this 

 group does not contain the largest cell bodies at this level. 

 They should be found in one of the lateral groups but, even for 

 the ventral group, without knowing anything of the peculiarities 

 of the gorilla in this respect, one would say that an average of 

 32 micra is small for an animal of this size. The two year old 

 gorilla in all probability has cell bodies smaller than the adult 

 but only slightly so. So far as has been investigated, animals 

 of the same species have in early life smaller cell-bodies of neu- 

 rones than the adult. In the neurological laboratory with which 

 the author has been connected this has been repeatedly proven 

 for the white rat, and Kaiser ('91) in another series of meas- 

 urements has proven it true for man and the white rabbit. 



Cavazzani ('97) determined the average mean diameters 

 of the cell-bodies found in the ganglia spinalia of a series of 

 mammals to which he also added the frog. His material was 

 fixed either with osmic acid or Muller's fluid and, in some 

 cases, his measurements were made from teased preparations 

 instead of from sections. His averages in each case are based 

 upon a large number of measurements (200 500) and should 

 therefore mean more for the ganglia than averages of a small 

 number of cell bodies especially chosen for their large size. His 

 measurements are made from the ganglia of the cervical, thoracic 

 and lumbar regions but are not confined to a given segment of 



