The Retina and Optic Ganglia in Decapods, especially in Astacus. 51 
Specimen A had 1155 corneal facets, and the number of retinal 
fibres must, tberefore, have been seven times as many or 8085. Spe- 
cimen B had 2375 facets, and its retinal fibres must have numbered 
16,625. 
In determining the number of fibres in the optic nerve it is 
important to bear in mind the two parts of the nerve: the dorsal 
portion, in which the separate fibres are distinguishable only v^ith 
difficulty, and the ventral part, in which the individual fibres are 
clearly outlined. As I shall show presently, it is to the ventral part 
that new fibres are added in the gradual growth of the optic nerve, 
the dorsal part being unaffected by these changes. In specimen 
the ventral part of the nerve contained by actual count 1291 fibres; 
the corresponding part of the nerve of specimen B contained 3430 
fibres. To form some idea of the number of fibres in the dorsal 
parts of the two nerves, I proceeded in the following manner. The 
difference in the number of facets in the two specimens, 2375 less 
1155, or 1220, shows the number of new ommatidia that have been 
added to the retina of specimen B in its growth beyond the condition 
represented by specimen A. The difference between the number of 
fibres in the two optic nerves, 3430 less 1291, or 2139, shows the 
number of new fibres added to the optic nerve in this growth. The 
ratio between the new ommatidia and the new optic nerve fibres is now 
easily found by dividing 2139 by 1220, the result being that the omma- 
tidia are to the fibres of the optic nerve as 1 to 1.75. Assuming this 
relation to be constant for the retina as a whole, the total number of 
fibres in the optic nerve can easily be estimated from the number 
of ommatidia. Thus in specimen A with 1150 ommatidia the total 
number of optic nerve fibres would be 1.75 times the number of 
ommatidia, or 2021 ; and in specimen B with 2375 ommatidia the total 
numher would be 4156. That these estimates come very near the 
truth is seen from the following relation. I have already stated 
that the dorsal part of the nerve does not receive new fibres in 
the further development of the optic apparatus, but the new fibres 
are added to the ventral part. The difference, then, between the 
total number of optic fibres and the number in the ventral part ought 
to be about the same in specimens A and B^ provided the estimates 
are nearly correct. Taking 3430, the number of fibres in the ventral 
part of 5’s nerve, from 4156, the estimated total for the same nerve, 
a difference of 726 is obtained. Applying the same Operation to 
the case of specimen B^ a difference of 730 appears. These results 
4 * 
