A loss of nerve cells in the cortex after encephalitis was reported already in the classical work by Nissl, Spielmeyer, and Spatz. A loss of nerve cells will become only noticeable if it amounts to at least 50%. But as such clear pictures are rarely found, estimations were always considered as doubtful and incorrect. Not only the number of cells is important in consideration of the morphological change in the cortex but also the size of cells. The development of a new apparatus made is possible to consider two structural parameters: the surface and the perimeter of cells. In 497 histological serial preparations obtained from 43 mouse brains we determined the number, the surface, and the perimeter of nerve cells. 39 animals were infected intracerebraly with yellow fever 17 D; 4 normal animals served as controls. Among the infected animals, 8 were treated with a mucopolysaccharide. The cells were counted within a determined area (standard unity); this area was taken from the angle between the curbura exterior and the sulcus anterior-posterior of the brain. There was a significant difference between the number of nerve cells in normal (278) and in infected (202) animals. The animals treated with mucopolysaccharide showed a normal quantity of nerve cells but surface and perimeter corresponded to the data of the infected ones. The surface of normal animals was at 23.39, that of infected at 14.29. There was also a significant difference with regard to the cell perimeter: normal 14.97, infected 12.02. This means a shrinking of cells. The cell shrinkage revealed that the nerve cells were affected. The measurement of these three parameters presents new and exact statistical findings which enable a reconsideration of neurovirulence.