We have earlier, at rau hei, guessed (from the glyphs in G) that the Easter Islanders possibly had the following 'map' for the phases of Venus:

Maybe we ought to revise the table by changing the time for transformation from evening to morning star from 10 to 8 nights (equal to the observed period). Before explaining why this seems reasonable, we should once again look at the evidence which made us earlier decide in favour of 10:

glyph numbers are counted beyond Gb2-25
3
	Ga3-10	Ga3-11	Ga3-12	Ga3-13	Ga3-14	Ga3-15	Ga3-16
	261	262	263	264	265	266	267
4				Haú at Ga3-14 (cfr 314) is glyph number 265 beyond haú at Gb2-25. 265 = 5 * 53, and 53 presumably functions as a symbol of the dark (e.g. beyond 52 * 7 = 364 days).
	Ga3-17	Ga3-18	Ga3-19
	268	269	270
5			The contrast between tagata with a vero head in Ga3-10 and tagata with a ragi head in Ga3-20 is significant. 270 = 584 - 314
	Ga3-20	Ga3-21
	1	2

If the corrrect number of black nights is 8, then we should change Ga3-18 and Ga3-19 from black to the ordinary blue colour above.

Maitaki in Ga3-17 would then be the last of the blackmarked glyphs, and 268 is an important numerical sign in the rongorongo system. The double manu rere (in Ga2-18), could indicate the reappearance of Venus as morning star - two 'birds' are in reality only one. The 8 dark nights are divided by haú into 5 + 3.

The structure of 5 + 3 dark nights can be perceived also in the text of Q, though in opposite order (3 + 5):

*Qb5-19	*Qb5-20	*Qb5-21	*Qb5-22	*Qb5-23	*Qb5-24
1		2		3

*Qb5-25	*Qb5-26	*Qb5-27	*Qb5-28	*Qb5-29	*Qb5-30
4		5		6
*Qb5-31	*Qb5-32	*Qb5-33	*Qb5-34	*Qb5-35	*Qb5-36
7		8		day 1

4 peculiar looking tagata rere (redmarked) characterize the 5 days preceding the 1st day of the calendar cycle.

Given this new information, I would like to change my earlier opinion of what possibly could be the Easter Islanders' view of the length of the phases of Venus:

I have chosen to keep 260 days for the evening star period, which necessitates changing 270 into 268. And we should then notice that the ordinal number for maitaiki (Ga3-17) is 268, a fact which has impressed me - the last day before the morning star reappears is number 268.

Changing from 270 to 268, in turn, necessitates changing 314 into 316. And we should then remember that the possible Venus day Gb1-21 has ordinal number 315, which means rau hei in Gb1-22 could indicate the last day before the evening star will rise:

Gb1-20 (314)	Gb1-21	Gb1-22 (316)

I have chosen to add 2 days (= 10 - 8) to the black period between morning and evening star. 32 is a good 'multiplication' number, better than 30, and 284 can be imagined as the difference between 584 and 300. This decision of mine means the last day for seeing the morning star should be at Ga8-16:

Ga8-13	Ga8-14	Ga8-15	Ga8-16 (284)
Ga8-17	Ga8-18	Ga8-19	Ga8-20

8 * 16 (in Ga8-16) = 4 * 32 and haú has 10 'feathers'.

Finally, I do not suggest that Venus as morning star is located at the beginning of the calendar cycle in Q (at *Qb5-35). What I do suggest is that the black winter solstice days have been likened by the creator of the Q text to the black 8 days before the morning star appears. Beyond winter solstice sun will reappear 'like a morning star in the east'.

The number of days in Q is 368, which can have been meant to allude to 268 for the evening star.