The development of the group spirit, with the appropriate sentiment of attachment or devotion to the whole and therefore also to its parts, is the essence of the higher form of military discipline. There is a lower form of discipline which aims only at rendering each man perfectly subservient to his officers and trained to respond promptly and invariably, in precise, semi-mechanical, habitual fashion, to every word of command. But even the drill and the system of penalties and minute supervision, which are the means chosen to bring about this result, cannot fail to achieve certain effects on a higher moral and intellectual level than the mere formation of bodily habits of response. By rendering each soldier apt and exact in his response to commands, they enable each one to foresee the actions of his fellows in all ordinary circumstances, and therefore to rely upon that co-operation towards the common end, be it merely a turning movement on the drill ground or the winning of a battle, which is the essential aim and justification of all group life. The group spirit, involving knowledge of the group as such, some idea of the group, and some sentiment of devotion or attachment to the group, is then the essential condition of all developed collective life, and of all effective collective action; but it is by no means confined to highly developed human associations of a voluntary kind. Whether the group spirit is possessed in any degree by animal societies is a very difficult question. We certainly do not need to postulate it in order to account for the existence of more or less enduring associations of animals; just as we do not need to postulate it to account for the coming together of any fortuitous human mob. Even in such animal societies as those of the ants and bees, its presence, though often asserted, seems to be highly questionable. When we observe the division of labour that characterises the hive, how some bees ventilate, some build the comb, some feed the larvae and so on; and especially when we hear that the departure of a swarm from the hive is preceded by the explorations of a small number which seek a suitable place for the new home of the swarm and then guide it to the chosen spot, it seems difficult to deny that some idea of the community and its needs is present to the minds of its members. But we know so little as yet of the limits of purely instinctive behaviour (and by that I mean immediate reactions upon sense-perceptions determined by the innate constitution) that it would be rash to make any such inference. The same may be said of associations of birds or mammals, in which division of labour is frequently displayed; when, for example, it is found that one or more sentinels constantly keep watch while a flock or herd feeds or rests, as is reported of many gregarious species.