150 SECTIONAL ADDRESSES. 
the third. And this would mean, of course, that totemism has not had 
an origin. 
In many of the theories of totemism it is difficult to tell whether the 
author is thinking of the first or the second of the two possibilities 
mentioned above. Prof. Elliot Smith, however, definitely adopts the first. 
If I understand him he would regard everything all over the world that 
he calls totemism (and J am not sure what he would include in or exclude 
from that term) as being derived in comparatively recent times from 
Egypt, and from a particular system of beliefs and practices which 
were the product of the special historical developments of Egyptian 
civilisation. 
Sir James Frazer’s final theory of totemism is well known to you. It 
assumes that all existing forms of totemism are derived from one simple 
original form. In making an assumption of this kind Prof. Elliot Smith 
and Sir James Frazer agree, but their agreement goes no further. The 
particular form selected by Sir James Frazer is what he calls conceptional 
totemism, the belief that the foetus in a mother’s womb is derived from 
some food (animal or vegetable) that the mother has eaten. ‘The belief is 
known to exist in parts of Australia and Melanesia, and I should think that, 
if it were sought for, it might quite well be found in other regions from 
which it has not been recorded. This, then, on Sir James Frazer’s theory, 
gives us the historical origin of totemism. It is not clear whether he 
conceives this form of totemism to have come into existence only once 
at a particular time in a particular spot, or whether he conceives it as 
having come into existence in different regions at different times. In 
completion of this theory he offers us a psychological explanation of the 
belief which, for him, is the germ out of which all diverse forms of totemism 
developed. Man, not being aware of the physiological causes of impregna- 
tion, but being desirous of finding some explanation, was led to the — 
theory that food eaten by a woman and followed by sickness (the 
sickness of pregnancy) was the cause of the pregnancy, with which it was 
thus associated. 
I do not intend to offer you criticisms of these two theories of totemism. 
If criticism is to consist, as I think it always should in science, of a 
re-examination of the evidence adduced in favour of-a hypothesis, [ cannot 
see that any evidence has yet been offered for the historical reality of 
either of these hypothetical processes. Indeed, I find it impossible to 
imagine what real evidence of that kind could be discovered. 
For comparative sociology, totemism presents a different problem or 
series of problems. These may be described as being concerned with the 
nature and function of totemism. To elucidate the nature of totemism 
we have to show that it is a special form of a phenomenon much more 
widespread, and we must aim at demonstrating that it is a special instance 
of a phenomenon or at any rate of a tendency which is universal in human 
society. For this purpose we have to compare totemism with all other — 
related institutions in all cultures. 
From the outset of our inquiry, therefore, we cannot isolate totemism 
and deal with it as a separate thing. First of all totemism in any given 
culture is part of a more extensive system of beliefs and customs, and may © 
occupy a preponderant position in that system, as in many Australian 
