On the assumption that vaha mea glyphs stand at the beginning of periods, we then will find a quarter (91 days) from Ha12-5 to Hb2-23. Likewise Ha6-42 will together with Ha7-31 measure out a lunar double-month (59). But Ha7-31 must then stand at the end of the period, and so must Ha12-4 if we wish to reach 4 * 64 beyond Ha7-31.

We have to abandon the idea that vaha mea glyphs always stand at the beginning of a period. Instead, the pattern which fits the numbers above is that there is one vaha mea at the beginning and one at the end of a numerical period. The visual cues also connect the vaha mea glyphs pairwise.

The only odd pair is the twin pair (Ha12-4--5) where the order is reversed - vaha mea at the end comes before vaha mea at the beginning. We will look closer on this phenomenon in just a moment.

140 we recognize as an important number from earlier, 10 fortnights. 48 added to 432 is equal to 8 * 60 = 480, 24 'twenties'. Together with the 7 'twenties' in 140 we have 620 = 31 * 20. This pattern emerges because we no longer count each side separately. 648 glyphs on side b has changed into 648 - 73 + 316 = 891, and then there are 271 glyphs between Hb5-16 and Hb10-8 which have not been counted among the multiples of 'twenties'. These 271 glyphs deserve a special discussion. By the way, 891 = 9 * 99, and 2 * 648 - 891 = 5 * 9 * 9, i.e., 1296 = 9 * 12 * 12 (because 99 + 45 = 144).

As to the visual cues we can imagine a moon crescent forming the jaw in the pair Ha2-23 and Hb5-16. More normal vaha mea jaws are in the 2nd pairs on each side (Ha12-4--5 respectively Hb10-8 and Hb10-55). We can therefore guess that also the first pair (Ha6-42 and Ha7-31) have jaws referring to the moon, or at least not to the sun. The number patterns 59 and 140 evidently refer to the moon, while sun is reflected in 8 * 60, 8 * 32 and in ¼ of 364.

We should also notice that the twin sun pair on side a are thick, as if pregnant.