[59] Professor Gummere also gives this analysis of Karl Bücher's essay on "Labor and Rhythm" (_Arbeit und Rhythmus_, Leipzig, 1896): "Fatigue, which besets all work felt as work by reason of its continued application of purpose, vanished for primitive man, as it vanishes now for children, if the work was once freed from this stress of application and so turned to a kind of play. The dance itself is really hard work, exacting and violent; what makes it the favorite it is with savages as with children? Simply its automatic, regular, rhythmic character, the due repetition of a familiar movement which allows the mind to relax its attitude of constant purpose. The purpose and plan of work involve external sources and external ends; rhythm is instinctive, and springs from the organic nature of man; it is no invention. The song that one sings while at work is not something fitted to the work, but comes from movements of the body in the specific acts of labor; and this applies not only to the rhythm, but even to the words. So it was in the festal dance.... That poetry and music were always combined by early man, and, along with labor, made up the primitive three-in-one, an organic whole, labor being the basal fact, with rhythm as an element common to the three; and that not harmony or pitch, but this overmastering and pervasive rhythm, exact, definite, was the main factor of early song,--these are conclusions for which Bücher offers ample and convincing evidence." (_Ib._ pp. 108, 109.) APPENDIX TABLE ILLUSTRATING THE HISTORY OF THE HEROIC COUPLET The following table is designed, in the first place, to illustrate the history of the decasyllabic couplet in English verse, by making possible a comparison of the characteristic details of its form in different periods; and, in the second place, to suggest a method by which, through the careful tabulation of facts, one may substantiate or correct general statements as to the qualities of verse. Any such statistics as are here presented must be accepted, if at all, with no little caution. The table is based on passages of one hundred lines each, believed to be fairly representative of the verse of the several authors. No passage of this length, however, can be known to be perfectly typical. A still greater difficulty is found in the necessarily subjective character of any such tabulation. Most lines of English decasyllabic verse can be read--with reference to the distribution of accents and pauses--in more than one way. It is unlikely, therefore, that any two readers would reach the same results in trying to form a table of this kind. The _absolute_ validity of the figures is therefore doubtful; but on condition that they all have been computed by the same person, consistently with a single standard of judgment, their relative validity, for purposes of _comparison_, may be fairly assumed.