This historic book may have numerous typos and missing text. Purchasers can download a free scanned copy of the original book (without typos) from the publisher. Not indexed. Not illustrated. 1911 Excerpt: ...irlandais," Revue celtique, 1904, xxv. 136). He accepts Roscher's theory of the sidereal month having furnished the basis for such 9-day periods as are found among the Celts, and argues further, that subsequently, the sidereal, having been abandoned for the synodic month, the 9-day periods became artificial units, independent of any connection with the moon. But so strange a transition as that from the sidereal to the synodic month cannot be supported by any Celtic evidence and has no analogy among other peoples. The number 9, it is worth while noting, enjoyed a high degree of sanctity among various Indo-Germanic races (O. Schrader, Reallexikon, 970 sq.). It is possible, however, to conjecture that the 9-day period arose, originally, as a multiple of a 3-day cycle in earlier use. 45R. E. Dennett, At the Back of the Black Man's Mind, London, 1906. pp. 214, 364. Cf. also, Waitz, op. cit., ii. 224 (Old Kalabar). northern Abyssinians. Every month contains four weeks, of which the first two are those of the increasing moon and the last two those of the decreasing moon. People say, however, that in reality the month has only three weeks and six days--and so it is (und das ist wahr).4s This Abyssinian evidence may be taken as making it reasonably certain that the Roman nundinal period (supra), the only 8-day week found in archaic culture, arose from the quartering of the lunation, since the length of the latter (29 days) may be calendarized in quarters, either of seven or of eight days. The Roman month was originally lunar and at all periods was divided by the real or (later) by the imaginary phases of the moon.47 In historical times, however, the nundinal period, like the Jewish seven-day week, was independent of the moon, running unfettered from month to mo...