Whoever has had patience to read through the material collected in the previous chapter will now no doubt be clear as to the process by which the cycle of months arose. The necessity was felt of distinguishing the months, of marking them. After the fashion of primitive man this was done, not by means of an abstract enumeration, but by some concrete reference. But the relation to a solitary historical event, by which rather more highly civilised peoples denote the years, can hardly, or only in isolated instances, be applied to the month: for the life of primitive peoples is very monotonous, and is not so rich in events which make an impression upon the mind that one of these will occur in every month, and even supposing that such events could be found, the months in a human life are too numerous for it to be possible to keep a series of this nature in mind. A second circumstance also proved decisive. The moon, whose phases always recur with regularity, served better than anything else to determine the date of any future event within a shorter period. The primitive peoples, with their undeveloped faculty of counting, could in this fashion numerically determine only a couple of months before or after the time of the moon that was then visible in the heavens. This is what we must understand by the statement made for the western tribe of the Torres Straits, viz. that they had no division of the year into months or days and never numbered the years, in view of the following statement that they commonly counted time in ‘suns’, i. e. days, and ‘moons’, i. e. months[806]. That is, they numbered two or three months, but had no series of months. The same initial stage is found also on the Australian continent. The natives of Central Australia reckon time by moon-phases, moons, and in the case of a longer period by seasons[807]. The Kakadu of Northern Territory reckon in moons and seasons, otherwise everything is more or less vague with the exception of the present and the immediate past and future[808]. Primitive man does not get very far in this fashion. In accordance with his custom and his whole habits of thought he must have some concrete factor to enable him to conceive of the different moons. This is found in the fact that the moon covers a part of the natural year. Herein lies a connexion which constantly recurs. The moons were therefore distinguished and named with reference to the phenomena of the natural year, to the phases of nature and to the occupations, labours, and conditions determined by them, and further to the risings of the stars. Within the series of from twelve to thirteen moons the month was determined by these means. Or, expressed somewhat differently, seasons and moons were mutually connected. Originally this grouping together of the months was only incidental. The original state of affairs is well illustrated by the detailed description given by Codrington for the Melanesians:--