480 THE NERVOUS SYSTEM. 
consisting of two longitudinal thickenings of the epiblast, it 
undoubtedly undergoes segmentation, and is subsequently 
found to consist of an integumental layer of columnar cells, 
and of deeper layers of irregularly polyhedral cells; but these 
deeper cells do not become directly converted into the ganglia ; 
the latter first appear as small groups of cells imbedded in the 
deeper layer, and only form a very small part of the primitive 
band. Compare Kowalevski’s figures [97, Pl. X., Figs. 31 to 39]. 
Possibly each ganglion originates from a single cell by rapid cell 
division, or from a small group of cells; but Iam unable to 
state anything definitely as to the origin of the primitive 
ganglia. All the cells in each ganglion in my preparations are 
precisely similar, and are very definitely separated from the 
larger surrounding cells. It is only subsequently, when the 
ganglia have attained considerable dimensions, that they 
assume the form of a segmented cord on either side of a 
very narrow median fissure. 
Quite recently Viallanes has described the manner in which 
the ganglia are developed in the embryo of Mantis (MM. religrosa) 
[190]. He says: ‘At first the primitive band is a simple 
thickening of the ectoderm—that is to say, a region in which 
the cells have become columnar and augmented in volume ; 
very soon these cells multiply and divide into two layers, a 
superficial dermato-genetic, and a deep ganglio-genetic layer. 
At a period which varies according to the region examined, 
the dermato-genetic layer separates from the ganglio-genetic, 
and becomes the hypoderm. Viallanes further states that 
the ganglio-genetic cells give rise to ganglion cells by division, 
and that the fibrillar or punctate substance is developed 
between the latter. Wheeler [191], who follows Viallanes 
very closely, describing the origin of the ganglia in Xiphidium 
ensiferum, one of the American Locustide, speaks of the 
ganglio-genetic cells as ‘large clear neuroblasts, scattered 
beneath the superficial cells of the blastoderm.’ Such may be 
the origin of the nervous elements ; but the appearances with 
which I am familiar are very unlike those figured by Wheeler. 
The appearances represented in Viallanes’ figures closely 
