Before proceeding to write in the glyph dictionary about Haga Takaúre it is necessary to look also in the other direction. There are takaure signs in glyph lines b2-4 and they need to be integrated into the scheme:

Gb2-34	Gb3-1	Gb3-4	Gb3-5	Gb4-4	Ga4-2
1	3	6	7	36	268

Ga4-2 has 'eyes' in both directions, a sign of standing at a border. Takaure in Gb3-5 has only a 'future eye' and is inside an oval with vai characteristics - it is not meant to show a canoe.

Gb3-4 and Gb4-4 have what could be a 'moon eye' at left, and they both are associated with 4. Gb4-4 could mark 4 * 4 = 16 and the end of a moon season. Presumably I did the right thing when I classified this glyph as a variant of takaure, although at that time I still had not come to appreciate the value of numbers.

Counting from Gb2-34 to the end of side b there are 183 glyphs (half 366). The glyphs in lines a1-a3 sum up to 83, and therefore Ga4-2 will have ordinal number 183 + 83 + 2 = 268, maybe a number related to 168.

Counting from Gb4-4 to Ga4-2 we have 232 = 8 * 29 glyphs. Gb4-4 (with ordinal number significantly = 36 counted from Gb2-34) could measure a full 8 months cycle together with Ga4-2. But such cycles should be measured by 29.5, not by 29. Therefore we presumably should start from Gb4-1, the glyph beyond the special 'fire generator':

Gb3-25	Gb3-26	Gb3-27	Gb3-28	Gb3-29	Gb3-30	Gb4-1
Gb4-2	Gb4-3	Gb4-4	Gb4-5

It makes sense in a peculiar roundabout way. Gb4-1 initiates a season with sun hidden inside, alluded to by a pau sign in Gb3-30. On side b there are 242 glyphs. There are 91 glyph (the number of a quarter) in the first 3 glyph lines, leaving 242 - 91 = 151 glyphs for the 5 following lines.

Ga4-2 is glyph number 85 on side a. And then, when we add the glyphs from Gb4-1 up to and including Ga4-2 we reach 151 + 85 = 236 = 8 * 29.5 - remarkable!

What does it mean? Is there a 'sun-moon' quarter (both 84 and 91) initiating the text on side b, followed by 4 moon double-months?

Gb1-1	Gb1-2	Gb1-3	Gb1-4	Gb1-6	Gb1-7
Gb3-25	Gb3-26	Gb3-27	Gb3-28	Gb3-29	Gb3-30
86	87	88	89	90	91

If we count from Gb1-7 (which we have done earlier), then there are 84 glyphs from Gb1-7 to Gb3-30. The disorder created by 64 is eliminated by renewing the counting from Gb1-8, it appears:

236 + 236 + 84 = 472 + 84 = 16 * 29.5 + 3 * 28 = 556.

But then we have counted Ga1-1--Ga4-2 twice. 556 - 85 = 471.

The text in G contains 16 moon months (29.5) divided in the middle by a gap containing 3 'sun-moon' months (28), and with an overlap caused by another set of 3 such months (with the addition of 1):

	24		58		64		84
Ga1-1		Ga1-26		Ga4-2		Ga6-10		Gb1-6	Gb1-7
1		26		85		150		235	236
8 * 29.5

	82			234
Gb1-8		Gb3-30	Gb4-1		Ga4-2
1		84	1		236
3 * 28			8 * 29.5

The takaúre season is here mapped as the 2nd half of the moon calendar, pushed away from the 1st half by a special quarter after midsummer and thereby causing itself to intrude upon the 1st half of the moon calendar in a corresponding measure.

Is this a description of a change caused when the 'carapace' was relocated in time? 16 * 29.5 = 472 could have been a greater waxing moon measure. 12 * 29.5 = 354 is too short, and the gap in time between consecutive years could have been mended by counting well past the darkest time.

Anyhow, here we have a reason for why waxing moon was counted - as we several times have had reason to conclude - as 16, not the otherwise expected 15 for the full moon night. It also explains why odd numbers (e.g. 15) were regarded as unlucky - there must be an even number in order to evade a dangerous 'crack in time'.

Moon rules the dark half of the year and sun the light half. Yet, sun is reborn already at the beginning of the year, which will tend to push moon backwards to be reborn already at the beginning of the 2nd half of the year. Which in turn will coincide with a wish to cut off the head of the dangerous spring growing beast. The concept seems to be expressed in the calendar of the day according to Q, where there is an abrupt 'cutting off' of the flow of sun time:

 

					-
Qa5-51	Qa5-52	Qa5-53	Qa5-54	Qa5-55

The 'tail' in hakaturou does not continue but has been given a sign of 'ghosthood'. The creator has recorded a gap in time from noon to the beginning of the night. The structure is similar to the gap in moon time from midsummer to the beginning of the 4th quarter.

One could even imagine this structure has been used by the creator of the rongorongo sign system as a form for the open mouth in such glyphs as e.g. Qa5-51. The head of sun cannot be drawn as a circle, because there is a gap beyond zenith (at the top part of the head).

At the beginning of the 4th quarter sun and moon are joining:

 

	6		82			234
Gb1-1		Gb1-8		Gb3-30	Gb4-1		Ga4-2
1		1		84	1		236
7		3 * 28			8 * 29.5
91

Gb3-30 can therefore be regarded as a 'bay of rest' and we can tentatively identify the beginning of the 4th quarter as Hanga Hoonu.

The 'carapace of the turtle' ends at haga hônu, in the same way as the 'carapace of the fly' ends at haga takaúre, and the distance between them is 8 * 29.5 = 236 glyphs. This is the shorter distance. The longer distance must be 236 + 84 = 320 glyphs.

236 + 320 = 556 and the normal length of the year (365) has been manipulated (56 contra 65, and 500 contra 300):

Instead of 365 = 364 + 1 = 13 * 28 (a theoretical construct counting only the nights when moon is capable of shining) the true moon cycle has been used to measure out 16 * 29.5 = 472 nights.

Instead of having a gap with 1 night between the years this crucial odd one seems to have been moved in the G text to the beginning of side a:

Ga1-1

It is needed to complete the 472 night long moon cycle.