This leads us naturally on to the second great division of our subject: progress in the methods of co-operation between man and man in doing industrial work. For if man is a social animal his power to do his bit and his consequent happiness must be derived, in part at least, from his social environment. The lonely craftsman perfecting his art in the solitude of a one-man workshop does not correspond with our industrial ideal any more than the hermit or the monk corresponds with our general religious ideal. It was the great apostle of craftsmanship, William Morris, who best set forth the social ideal of industry in his immortal sentence: 'Fellowship is Life and lack of Fellowship is Death.' Our study of the workman, then, is not complete when we have seen him with his tools: we must see him also among his workmates. We must see industry not simply as a process of production but as a form of association; and we must realize that the association of human beings for the purpose of industrial work involves what is just as much a problem of government as their association in the great political community which we call the State. It is difficult to see the record of the progress of industrial government in clear perspective for the simple reason that the world is still so backward as regards the organization of this side of its common life. The theory and practice of industrial government is generations, even centuries, behind the theory and practice of politics. We are still accustomed in industry to attitudes of mind and methods of management which the political thought of the Western World has long since discarded as incompatible with its ideals. Two instances must suffice to illustrate this. It is constantly being said, both by employers and by politicians, and even by writers in sympathy with working-class aspirations, that all that the workman needs in his life is security. Give him work under decent conditions, runs the argument, with reasonable security of tenure and adequate guarantees against sickness, disablement and unemployment, and all will be well. This theory of what constitutes industrial welfare is, of course, when one thinks it out, some six centuries out of date. It embodies the ideal of the old feudal system, but without the personal tie between master and man which humanized the feudal relationship. Feudalism, as we saw in our study of political government, was a system of contract between the lord and the labourer by which the lord and master ran the risks, set on foot the enterprises (chiefly military), and enjoyed the spoils, incidental to mediaeval life, while the labourer stuck to his work and received security and protection in exchange. Feudalism broke down because it involved too irksome a dependence, because it was found to be incompatible with the personal independence which is the birthright of a modern man. So it is idle to expect that the ideal of security will