PRESIDENTIAL ADDRESS. 631 
scious and intentional adoption, but also, as Robertson Smith well said, ‘experi- 
ence shows that primitive religious beliefs are practically indestructible, except by 
the destruction of the race in which they are ingrained,’ All apostles of rew 
creeds have had to preach to and gain the adherence of societies which they could 
not hope to lead to a perfect way all at once or even in centuries of time, and all 
have had to take account of pre-existing habits of religious thought and actual 
expressions of religious feeling, and by accepting some compromise to modify these 
to their purpose. And if this be obviously true of those societies which such an 
apostle as Mohammed could influence directly and retain under some sort of 
personal control, what must we say of the societies to which the truth only came 
at second hand or by many more degrees removed from the original prophetic 
utterance? What of the remote or scattered folk to whom it came not at all till 
after a long interval, and then faint and confused as a reverberating echo? For 
these at least there was no possibility of such utter change as revelation working 
through the human agency of a magnetic personality may have effected elsewhere ; 
and of their belief and their practice much, perhaps the most, has remained 
primeval and local, and as the physical conditions of their life have prompted it 
from all time to be, and prompt still. 
All this stratification in religious belief and practice it is the function of 
Anthropology to investigate ; and thereby it may render no small service to religion 
itself by distinguishing accidental elements in ritual and observance which have 
persisted from systems worn out and abandoned. But while proclaiming that this 
investigation is not only legitimate but necessary, I wish to-day to utter a note of 
warning against a certain confusion of thought which is often manifested by the 
investigators in this particular field, and is apt to occasion unfortunate ethical 
consequences or, at the best, unnecessary scandal. It finds expression in the 
grouping of all the elements in belief, observance, and ritual, which have persisted 
from earlier systems to later, under one head as religious survivals, without due 
account being taken of very vital differences, both in their essential nature and in 
the history and reason of their persistence. The word ‘survival’ itself is per 
accidens not a very fortunate one. Though in the broad sense perfectly appro- 
priate to all things that persist, it has acquired in our modern speech, largely from 
its use in medical science, a certain particular connotation of opprobrious import. 
It suggests something which has lost its useful purpose, and is effete or even dead, 
persisting among living organisms usually to their detriment. Such is the 
sense in which many anthropologists seem to use the word in speaking of religious 
persistences without discriminating between divers kinds of these; and such, still 
more often, is the connotation which their readers attach to the word in this con- 
nection. Yet all religious persistences are not survivals in this pathological sense 
—nay, the class to which this connotation is suitable includes but a small propor- 
tion of the whole. It is to distinguishing these classes of survivals that I propose 
to address myself in the remainder of the time which is allotted to me to-day. 
In the first place there is a most numerous and important body of religious 
persistences which ought not to be called survivals at all, if that word be used, 
as it usually is, with a causative implication; that is to say, there are elements of 
belief and practice whose existence in actual cult is not necessarily due at all 
to the fact that they, or something very closely akin, existed in a previous cult. 
If religion is the expression of the instinctive desire of man to find an intelligible 
relation between his own nature and a nature which transcends its limitations, 
he appears unable to establish that relation by other than a very small and 
definite number of conceptions; and among certain races, and indeed in certain 
geographical areas, those conceptions seem not to vary over immense spaces 
of time and under successive dispensations. The just way to regard them, 
therefore, is as falling within categories of thought inevitably imposed on the 
human mind by its humanity and necessary conditions of any religious sense 
whatever. Man does not form these conceptions because his predecessors formed 
them, nor indeed because his contemporaries hold them; but because, as an 
individual limited by race and environment, he cannot otherwise satisfy his 
religious instinct. How important this class is, and how much it includes which 
