The Retina and~Optic Ganglia in Decapods, especìally in Astacus. 41 
elements. As one can see in Golgi preparations (Fig. 46), as well 
as in ordinaiy transverse sections (Fig. 42), this layer is completely 
penetrateci by the bundles of retinal fibres. 
The proximal nuclear layer (Fig. 46, 4) is precisely similar to 
the distai one except that it contains on the whole fewer cells. 
What has been said concerning the nature of the distai cells applies 
with equal force to those in this layer. These two nuclear layers, 
the distai and the proximal, bave the appearance of having been ori- 
ginally a single layer, which became divided by the formation of 
the fibrous layer. The fact that the fibrous layer contains almost 
no cells leads one to look for the elements from which it arose in 
the layers adj acent to it. Whether it is a joint product of the two 
nuclear layers or not is stili to my mind an open question, though 
the Position that it occupies inclines me strongly to the opinion that 
it is produced by these layers. 
The last layer in the ganglion is the layer of »Punktsubstanz« 
(Fig. 46, 5). To this the retinal fibres make their way and from it 
emerge the fibres that connect the first and second ganglia. The 
»Punktsubstanz« of this layer is not of uniform consistency, but is 
divided into small masses that are so regularly arranged that in 
transverse sections (PI. 2 Fig. 44) they bave somewhat the appea- 
rance of honey comb. Their arrangement and size are such that the 
question naturally suggests itself, are not these bodies the ganglionic 
representati ves of the retinal ommatidia? Viallanes (92a, pag. 391), 
who has studied them in Palinurus^ believes that they do correspond 
to ommatidia and Supports this opinion by the statement that each 
ganglionic body, which he calls by the very appropriate name of 
neurommatidium, contains seven fibres, the exact number given out 
by each ommatidium. Owiug to the unfavorable condition presented 
by the ganglion in the crayfish, I bave been unable to obtain decisive 
evidence for or against this opinion, and the only pertinent Obser- 
vation that I was able to make was in regard to th-e total number 
of retinal and ganglionic elements. In an eye with 1798 ommatidia 
the estimated number of neurommatidia was 1833. This, though 
not a perfect coincidence, is near enough to lead me to believe that 
Viallanes is correct in bis opinion that the neurommatidia agree 
in number with the ommatidia. 
Another question of importance concerning the «Punktsubstanz« 
has to do with the fibres that enter and leave it. Viallanes (92a, 
pag. 392) States that the retinal fibres pass directly through the first 
