Let us take the example given of the respect for human life as a crucial test. Admitting that religions have taught that to take life was a sinful act, one might well interpose with the query as to whether it was ever necessary to teach man that homicide within certain limits was a wrong thing. Pre-evolutionary sociology, which sometimes taught that man originally led an existence in which his hand was against every other man, and who, therefore, fought the battle of life strictly off his own bat, may have favoured that assumption. But that we now know is quite wrong. We know that man slowly emerged from a pre-human gregarious stage, and that in all group life there is an organic restraint on mutual slaughter. The essential condition of group life is that the nature of the individual shall be normally devoid of the desire for the indiscriminate slaughter of his fellows. And if that is true of animals, it is certainly true of man. Primitive human society does not and cannot represent a group of beings each of whom must be restrained by direct coercion from murdering the other. In this case, therefore, we have to reckon with both biological and sociological forces, and I do not see that it needs more than this to explain all there is to explain. Human life is always associated life, and this means not alone a basis of mutual forbearance and co-operation, but a development of the sympathetic feelings which tends to increase as society develops, they being, as a matter of fact, the conditions of its growth. And whatever competition existed between tribes would still further emphasize the value of those feelings that led to effective co-operation. The question, then, whether the anti-homicidal feeling is at all dependent upon religion is answered in the negative by the fact that it ante-dates what we may term the era of conscious social organization. That of whether religion strengthens this feeling still remains, although even that has been answered by implication. And the first thing to be noted here is that whatever may be the value of the superstitious safeguard against homicide it certainly has no value as against people outside the tribe. In fact, when a savage desires to kill an enemy he finds in superstition a fancied source of strength, and often of encouragement. Westermarck points out that "savages carefully distinguish between an act of homicide committed in their own community and one where the victim is a stranger. Whilst the former is under ordinary circumstances disapproved of, the latter is in most cases allowed and often regarded as praiseworthy." And Frazer himself points out that the belief in immortality plays no small part in encouraging war among primitive peoples,[19] while if we add the facts of the killing of children, of old men and women, and wives, together with the practice of human sacrifice, we shall see little cause to attribute the development of the feeling against homicide to religious beliefs.