Remarkably I found the 18 glyph discrepancy to have its correspondence in the appearances of the glyphs. To explain we should first once again read the following:

... In Aa5-7 the first complete viri of the solar year appears:

1	2	3	4	5
6	7	8	9	10

11	12	13	14	15	16

17	18	19	20	21	22	23

24	25	26	27

28	29	30	31	32	33

34	35	36	37	38

39	40	41	42	43	44	45	46

I have numbered starting from Aa4-63 because that is where the internal long parallel with side b begins (at Ab5-1). Redmarked are 26 and 46 because 1334 = (26 + 20) * 29.

The old bird at 46 is GD23 which I earlier have identified as meaning 'ending' (= reversal and new beginning).

Evidently viri is here outside - given that Aa4-63 is inside - the measure of 26.

However, another indicator says that Aa4-63 is outside (and viri no. 26)  - because counting from the beginning of side a (from Aa1-1) we have 314 (= 100*π) as Aa4-64:

Another 'proof' of viri being inside is obtained by counting from the beginning of line a5:

Aa5-1	Aa5-2	Aa5-3	Aa5-4	Aa5-5	Aa5-6	Aa5-7
0	1	2	3	4	5	6

Aa5-1 is GD42 which I earlier have identified as a marker of beginning. Excluding Aa5-1 viri is no. 6, the solar number. Adding this 6 to the earlier 20 we receive the wished for 26.

I now suggest that Aa5-17 may contain the missing second part of the viri that is embedded in Aa1-1:

Aa5-17 is an aberrant form of GD26 with the 'head' as a separate part, the only such in Tahua (in the same way as Aa1-1 is an aberrant unique form of GD42). Furthermore, we recognize hakaturu, the point of turnaround, as Aa5-18 (no. 350 from Aa1-1):

Moving 18 glyphs backwards from Aa5-7 we arrive at another GD42, the one who announces a break in time-space:

Aa4-63	Aa4-64	Aa4-65	Aa4-66	Aa4-67
Aa4-68	Aa4-69	Aa4-70	Aa4-71	Aa4-72

From the GD42 bird in Aa4-71 to the one in Aa5-1 there are 12 glyphs, possibly a sign denoting that half the year has passed. In Aa4-72 we can see the visualized expression of that. A break is evident between the left and right part, reminding us about Aa1-15:

72 (as in Aa4-72) is 360 / 5, or rather: 2 * 36.

Whatever the reason may have been for placing the 'break' glyph (Aa4-72) 18 glyphs before viri (Aa5-7), and the count here includes both the 'clew' and the 'koti', we can see that there is a correspondence between the glyphs (how they look and where they are located) and the numbers of glyphs.

18 emerges from 16 (twice 8) and 29, and these numbers emerge from the 4 viri and the 670 + 664 = 46 * 29 glyphs on the two sides of Tahua.

The complex relations are probably not figments of my imagination, there is too much evidence for that. We can confidently use the results as a foundation for further investigations.

To make the structure easier to remember we need a picture:

I here decided to investigate and list all the 'break' glyphs in Tahua, it might be useful later on:

The definition I used was that the glyph should have a horizontal 'break'. Vertical breaks occur too, for example in Aa2-3 and Ab4-18, but that is another matter:

Possibly there is a 3 + 1 pattern in the distribution of the glyphs in the table above:

Yesterday I updated the 'definition' of GD33 in the glyph catalogue, i.e. which visual cues in the glyph should be used for classifying a glyph as belonging to GD33.

The definition of GD33 begins like this:

Accordingly, only Ab1-1 seems to be 'unblemished' by marks. My idea to turn around Aa8-26 counterclockwise 90º seems ill advised. After all, we have not only L1-33 but also to Eb6-24 to consider:

How big a glyph is probably depends on how great a season it refers to. Therefore, Ab7-26 is 'greater' than Ab1-1. That we have understood already from the numerical exercises: 29 * 26 is greater than 29 * 20, and both Aa8-26 and Ab7-26 belong to the greater period, not to the period which incorporates Ab1-1:

	60 = 3 * 20		520 = 26 * 20
Aa8-26		Ab1-1		Ab7-26
20 * 29

If we do not turn the viri glyphs away from how they are drawn, the top presumably represent a later stage than the bottom - growing things emerge from the ground slowly reaching higher levels.

In Aa8-26 (the end of the greater period) the bottom therefore may represent the 3rd quarter of the solar year and the top the 4th and last quarter, which is abruptly cut short where the 5 dark nights intrude. That would explain why the top part is thinner than the bottom part: What is represented is the force of the solar power, I guess.

In Ab7-26 - 7 lines away - the reverse is seen. The top part is thicker than the bottom part. Accordingly I suspect that here the top part represents the 2nd quarter and the bottom part the 1st quarter of the solar year.

Between Aa8-26 and Ab7-26 we ought to find 'winter', while the greater cycle (26 * 29 = 754 glyphs long), from Ab7-26 to (and including) Aa8-26 presumably represents 'summer'.

16 stages into the greater cycle we find Aa5-7, which I earlier have described as 'deformed' (... Aa5-7 and Aa8-26 are drawn as if deformed ...):

	482 = 16 * 29 + 18		272 = 10 * 29 - 18
Ab7-26		Aa5-7		Aa8-26
26 * 29

Neither the top part nor the bottom part in Aa5-7 is thick, and the glyph is slightly bent forward, as if it was old. The fire is about to fade away - after 'noon' sun is descending.

16 * 29 = 464.

If there are 16 parts in the 1st half-year (from solstice to solstice), then these 16 should be divided into 4 groups with 4 parts in each. 464 / 4 = 116.

We can try by counting from Aa1-1:

116

Aa2-26

Metoro identified Aa2-26 as Rongo: 'ka rogo ia'. Maybe he saw this person's right strangely formed 'ear' as a symbol for the 1st half year. We remember the similar symbol on top of the right staff of the sun god in Tiahuanaco:

In the month signs we find the opposite symbol at the bottom of the 'eye stream':

A pair of oppositely oriented half-circles were used as symbols in ancient Egypt too:

Maybe two joined half-circles are to be read in GD33? The 'knee' at the joint could be autumn equinox and this joint could be the reason why Metoro often said viri at GD33 glyphs: