in some prospective audience for his achievements—there can be no doubt that literature, whether its form be verse or prose, whether its spirit be intensely individual or social, cannot live apart from some kind of sympathetic group. Hence the gradual extension of social sympathy is a leading feature of literary development. Sensations and emotions, moral and intellectual ideas, widen in their range as smaller groups merge into larger; and this progressive merger underlies the development of institutions and language, and is largely a maker of myth, the interlacing of group traditions (easily illustrated by early Arab history) creating eponymous myths of every variety.
But it is not to be supposed that the social groups, or their progressive merger into widening circles, admit of exact definition. We cannot select any exact point and say, "Here the clan breaks up and the city succeeds." We cannot deny the existence of clan associations in days when the leading features of clan life have been obliterated—nay, it is often only because such survivals have taken place, owing to the progress of different parts of society at different rates, that we can recover the past at all. We can draw no hard-and-fast lines limiting the idea of "clan" absolutely, or denying its concrete existence save where our ideal definition is exactly realised. And the reason of this apparent vagueness is the best of all possible reasons—viz., the absence of any such definite lines in social phenomena not abstractly sketched on paper, but as living and moving realities. The jurist does not attempt to minutely define the Roman familia, or give the exact dates at which the marked features of that social unit faded away. The economist, with all his masses of statistical information, cannot distinguish his capitalist and labouring groups save by broad distinctions which in concrete life insensibly fade into each