IN THE SHETLANDS 103 
would, I believe, have been sufficient ; for it must be 
remembered that man’s ancestry leads up, through 
the semi-human, to the primeval savage, and it is 
amongst the lowest tribes of existing savages that 
the tendency last indicated is most noticeable. In 
regard to this, one should read Tylor, as likewise 
Clodd, concerning the probable effect that dreams 
have had in producing the idea of a soul. From 
the dream figure to that of our waking mind’s eye 
there is but a step; and as animals dream, we may 
suppose that they likewise see mentally. 
This seems clear, that wherever the visualising 
faculty—to give it a name—produced the image of 
anything, it would be mistaken for that thing if 
reason did not convince to the contrary. In animals, 
reason is weaker than with us, but that the power of 
mental vision, within the narrower range of their ex- 
perience, is weaker also, I can see no reason to con 
clude. Rather, I think, it is likely to be the other 
way, and this should make it an easier matter for 
a guillemot than for a negro to see, or seem to see, an 
absent relative. But possibly this vivid conjuring up 
of the mere outward form of anything may not be 
required in order to induce the belief of its being 
there. The negro, perhaps, rather feels the presence of 
his mother than thinks that he actually sees her; and 
might not this effect, also, be produced through a strong 
association of ideas? If so, this is all that would have 
been necessary to give man that belief in his immortal 
destinies which, upon the whole, we find him with. 
