856 THE AMERICAN NATURALIST. (VoL. XXXII. 
must be determined in part by the impulses they receive from 
the retina, just as certain joints, for example, have been pro- 
duced by repeated stresses and movements of a particular 
kind. But when such joints are once established, they permit 
only those kinds of movements that were most instrumental in 
producing them, whatever may be the nature of the stimulus 
that initiates these movements. It is, therefore, probable, when 
certain adjustments have once been established in the optic 
ganglion through repeated retinal stimulation, that any stimula- 
tion of such a collection of fibrils might call forth the particular 
series of activities necessitated by-such an adjustment, whether 
the initiatory impulse comes from within or from without. In 
other words, the existence of a second visual centre, or optic 
ganglion, having a structure similar to that of the retina, may 
be considered as essential in establishing for past experiences - 
latent records, which can be brought again into activity through 
other stimuli than those that originally produced them. We > 
are thus provided with a physical basis for the explanation of 
hallucinations and for certain phenomena of visual memory. 
It is clear from known anatomical relations that the visual 
impulses do not cease at the optic ganglion, but are transmitted 
to other centres in the cerebral hemispheres, provided such 
hemispheres are present. While there is no animal in which 
the optic ganglion is not united with the anterior part of the 
brain, it is apparently only in the higher arthropods and in the 
vertebrates that it is united with a definite part of the cerebral 
lobes. In Limulus and scorpions, where I have made a special 
study of these conditions, the cerebral visual centres do not 
resemble in any way the optic ganglia or the retina, but their 
whole appearance indicates that they consist of masses of cells 
and fibres that serve to bring the nervous impulses received 
from the eyes and optic ganglia into relation with various parts 
of the body. 
We are thus led to conclude from a comparison of widely 
different types of animals that the visual apparatus consists of 
three principal parts, which phylogenetically are developed m 
the order named, viz.: the retina, or receiving centre; the optic 
ganglion, or recording centre; and the cerebral portion, or the 
