Crip, 9) 
short, and is in most insects cut up into stages, the experience 
gained in one of which would be of little use in a following 
stage, where the creature is provided with a different set of 
structures; and in point of fact we see, from observation, 
that the butterfly, directly after it has come into possession 
of its beautiful compound-eyes, knows well some of those 
properties of space and matter that we learn only from expe- 
rience. So, again, with size; we may be sure that the insect 
does not infer this by a mental process: and it cannot 
associate tactile experiences with its ocular impressions, as we 
do, for not only must any tactile powers it may possess by 
inheriting its ancestors’ associations be of a very specialised 
and limited nature owing to its hard external covering, but 
its life is too limited to allow it to acquire such associations 
individually. Its eyes, too, are fixed and immovable, directed 
mainly one to one side and one to the other, so that it 
receives two very distinct sets of impressions from the two 
eyes. 
Under these circumstances it is clear that flat-picture vision 
is not a satisfactory function to assign to the compound eye 
of the insect. And even from this there is a most important 
deduction to be made. As yet we have not considered the 
mosaic part of the question. The picture formed on our 
retina, if defective in various other ways, is, at any rate, 
continuous; but in the insect, on the Mullerian theory, it is 
certainly broken up, and must be on this account less 
valuable. The perfection of a mosaic picture depends largely 
on the amount of cement there is in it in proportion to the 
marble, and where the seams of cement are coarse, the per- 
fection of the picture is greatly marred. The supporters of 
the mosaic theory state that the perfection of the mosaic 
picture in the insect’s eye depends on the number and the size 
(or coarseness, as entomologists say), of the facets; these 
vary excessively, and probably would be found in many 
insects to greatly reduce, if not altogether destroy, the value 
of the picture, which would indeed at best be, not a picture, 
but pieces of a picture separated by intervals of blindness. 
I am not aware whether any attempt has been made to 
calculate what proportion the cement would bear to the 
