Tagaroa uri is the 4th month in the Easter Island calendar for the year. With the calendar in E having 24 periods, we can be fairly sure that the Keiti calendar has two periods for each month. The 4th month can therefore be expected to be covered by periods 8-9, not to be located in period 6.

This inconsistency cannot be ignored. Maybe the Keiti calendar does not begin with July (He Anakena)? Spring equinox probably is located in period 5, and therefore periods 4-5 ought to be coordinated with September (Hora nui):

4
	Eb2-25	Eb2-26	Eb2-27
	manu ragi rima	manu ragi	kiore henua
5
	Eb3-1	Eb3-2	Eb3-3
	koia ra	te maitaki	te henua
	Eb3-4	Eb3-5	Eb3-6
	ihe vage Rei	oho mai kia iha	ki te henua kiore

Possibly periods 1-3 are to be considered as belonging to the end of the previous calendar cycle, i.e., the calendar of E could begin with the spring equinox month Hora nui.

Another alternative is that the preceding Hora iti is covered by a single period in the Keiti calendar - otherwise He Anakena would not have 2 periods. If this is true, then the (presumed) ancient model with only 10 months may have been transformed into a 12 month calendar by dividing an ancient Hora month into two parts, Hora iti and Hora nui, where iti ('little') got only half a month. Similarly Vaitu nui and Vaitu potu once may have been just one month, Vaitu (and Vaitu potu may have got only half a month).

With the suggested explanation another problem is, though, arriving: Why 24 periods if there are only 22 half-months in the calendar? The difference, 2 periods, can be motivated as being the same difference as between 26 and 24 in the kuhane stations - 2 extracalendrical stations. From Barthel 2 the following information can be coordinated:

Counting the glyphs in periods 1-2 in the Keiti calendar we reach 20 + 6 = 26, possibly a clue that there are 26 half-months in the calendar. Given that the calendar begins at new year with He Anakena, it would be suitable to set the number of glyphs in the first two periods equal to 26.