There are no ragi glyphs in the Mamari moon calendar and neither are there any ragi in the calendars for the week (in H and P). In the 2nd, 3rd and 5th periods of the G and K calendars for the year, however, we find signs of ragi. We begin, though, from the beginning:

I intended to investigate whether the K calendar was using a lunar perspective in contrast to a solar perspective in G, but already at the beginning difficulties were found:

In Ka3-17 the beak suggests moon, which was noted at niu in the glyph dictionary:

In the Translation part a more detailed discussion describes the arguments:

The parallel Ga3-1 is not manu rere but manu kake. Its assymmetry resembles that in Cb5-9:

Cb4-20 resembles the 'fire-generator' in e.g. Ka3-14:

Before Cb4-17 there is an 'end-of-period' glyph and beyond Cb5-21 a new glyph line begins:

The strange 'hook' in Cb6-1 is similar to the 'hook' in Ka3-14, possibly not a coincidence. Beyond Ka3-14 the 1st period of the calendar begins and Cb6-1 is the first glyph in a new line - 'hook' at left (past) is reasonable.

Reading glyphs together with their ordinal numbers seems to be the only secure way to read rongorongo texts. The 'eating mouth' (refilling station) has sun characteristics (a double rim), while the full cycle in Cb5-18 has only one rim (i.e. presumably means moon).

Beyond the final Cb5-20 with hua and ariki a ghostlike king remains (Cb5-21). At Cb5-10 (the halfway station) nuku with maro (3) is followed by ariki and a sign of death (ihe tau) - somewhat resembling Aa1-11:

Logically then, Cb5-9 may have a meaning similar to that in Aa1-10:

Is this not a clue to why the left head in manu kake is drawn smaller? Only one 'eye' in Aa1-10 - together with 'knee' at right - indicates 'past the center' yet not fully 'cooked', i.e. a curious 'half-cooked' character (still alive and hard to kill like a turtle).

Cb4-17--18 resemble the beginning of the 2nd month in the Mamari calendar:

Their 'greater' shapes, together with ordinal number 18, suggest the meaning is sun rather than moon, though.

Maybe in Aa1-10 we also should read two 'sun canoes', one dangling and newly released and one still 'inhabiting' the arm?

Is the 'canoe' in Aa1-2 a symbol for the sun or for the moon? Its ends are uplifted as in Cb4-17--18, yet it is slim like Ca9-1--2:

I think the 28 glyphs, Cb4-17--Cb5-21, are to be read as two subgroups, where the first subgroup ends with the last glyph in line b4:

I wrote 'subgroups' (not 'groups') because they must belong together - the number of glyphs otherwise would become odd. 7 + 21 = 28.

¼ + ¾ = 1, a pattern we recognize as '1 + 3'.

21, 42 and 84 are 1, 2 and 4. Glyph number 63 (= 3 * 21 = 9 * 7) in the K calendar, how does it look?

The impression is that honu at last is beeing 'fully cooked'. There is a balance between fully alive (mago) and fully dead (honu). Period 18 in the K calendar is located beyond autumn equinox we have read earlier, yet the seasons as described in the calendars are rather fluid. 18 - 6 = 12.

15 + 1 = 16 and 23 + 1 = 24, i.e. 2 * 8 respectively 3 * 8.

21, 42, 63 and 84 = 7 * (3, 6, 9 respectively 12).

The pattern '1 + 3' implies that the 'odd one' (1) has a 'ghostlike' character, at least judging from Kb2-4. We should look again for confirmation in line 18:

Period 18 corresponds to period 20 in G and the 'end-of-period' in G has ordinal number 29, beyond which we have:

Although G and K do not seem to walk hand in hand - cfr e.g. the 'solar fist' at right in Ga6-1 while it is at left in Kb2-103 - a new season is beginning in both calendars.

The subgroup '1' has in the 'twins' Cb4-22--23 'spooky' hands (3 together):

The number of eyes in Cb4-22--23 is also 3 (+ a special one at left in Cb4-23). The central 'story' evidently is:

Leaving in a canoe, making new fire, flying forward.

The figure in the canoe (in Cb4-19) appears again as number 10 in the 2nd subgroup ('3'), or as number 14 counted from Cb4-19: