5. The Pleiades cannot be observed at all times. Makemson says that 'thirty or forty days' after their disappearance 'in the rays of the setting Sun toward the end of April' the Pleiades 'are visible on the eastern horizon just before dawn':

It is not clear if also on Easter Island they defined periods Matariki i nika and Matariki i raro by observations in the evening, or in the early morning. In the latter case the table above is wrong. Likewise, of course, if they did not use the concept at all.

If they watched the Pleiades both in the morning and in the evening, it is conceivable that they had a new year which began according to evening observations and a ritual calendar which began according to morning observations.

The rays from the sun 'kills' a part of the cycle of the constellation. The stars are there, but cannot be seen. It is very much like when the new moon is still dark because it is in the sight line between the observer and the sun, bathing in 'vaiora a Tane'.

Otherwise a star or constellation can be observed during half a year (given a fixed time in the night for the observations), from its appearance in the east to its disappearance in the west. It takes ca 2 quarters (182 days) to travel 180º.

Once there were 13 constellations in the zodiac, but one was eliminated (the Serpent). If we count 12 / 13 of 182 days we get 168, an important number in the rongorongo texts. Maybe the period of invisibility was regulated to be 182 - 168 = 14 days. Twice that comes not far from Makemson's estimate 30 - 40 days.

Another alternative would be to stipulate 32 days' invisibility:

Te Pou (Sirius) is located at 9 * 29.5 = 265.5 (or rather day number 266) and 100 days later the solar year is ending at 365.25 (or rather day number 366). The redmarked alternative with 32 days of invisibility and 166 days of visibility is to prefer because of the harmony. And the ancient Egyptians associated the reappearance of Sirius with new year.