K calendar
period no.		number of glyphs
1, 2, 3	3	7+3+4 = 14
4, 5, 6, 7, 8, 9, 10, 11, 12, 13	10	3+2+3+2+2+2+2+3+2+3 = 24
14, 15, 16	3	2+3+9 = 14
		4
17, 18, 19, 20, 21, 22, 23, 24, 25, 26	10	3+6+4+4+3+6+2+2+3+3 = 36
27, 28, 29	3	5+4+5 = 14
30	1	4

The 30th period, as we remember, has a different end glyph (Kb4-18) and should not be counted in the same way as the 29 first periods:

The sequence of 4 black glyphs at the end of the 16th period has a similar glyph (Kb1-13) - though oriented the other way:

Ga5-10 is located at the beginning of period 18, while in period 16 of K there are 2 glyphs parallel to period number 16 in G and 3 glyphs parallel to period 17 in G, before Kb1-12 arrives as glyph number 6 in period 16.

2 + 3 + 5 + *2? + 1 = 13?

If Kb1-101 not is parallel to Ga5-16, but a different kind of glyph, there will be 28 (instead of 29) regular periods in K. Period no. 16 will then stretch 3 glyphs further on (covering what so far has been described as period 17).

The beautiful pattern with 36 glyphs in 10 periods (17-26) would then be destroyed, however, and therefore the hypothesis must be that Kb1-101 is a glyph which is parallel to Ga5-16 - first of all in its location but preferably also in its type.

On the other hand, the reconstructed ordinal number in the line for Kb1-101 (i.e. *19) seems to be wrong. *18 would be a better choice.

The situation - it may be argued - can be saved by suggesting that Kb1-101 has a reconstructed ordinal number *18 and that Kb1-101 is parallel in position and glyph type with Ga5-16 and that its colour is 'black'.

However, as a consequence the glyph number for period 16 in K will then be *12 (instead of 13?) and we will have 2+3+9 = 13 instead of 14 for periods 14, 15 and the beginning of period 16. That cannot be an acceptable solution, because 14 + 24 + 13 = 51.

Another possibility is to count only 2+3 = 5 glyphs (parallel to periods 16-17 in G), 14 + 24 + 10 = 48:

K calendar
period no.		number of glyphs
1, 2, 3	3	7+3+4 = 14	48
4, 5, 6, 7, 8, 9, 10, 11, 12, 13	10	3+2+3+2+2+2+2+3+2+3 = 24
14, 15, 16 (start)	3	2+3+5 = 10
16 (end)		7
17, 18, 19, 20, 21, 22, 23, 24, 25, 26	10	3+6+4+4+3+6+2+2+3+3 = 36	50
27, 28, 29	3	5+4+5 = 14
30		4

The structure 3+10+3+10+3 = 29 periods is intact, but we have now reduced the total number of glyphs to 98+11 =109, which indicates that the black 'periods' are not yet described properly.

We return to the glyph dictionary:

Green, red, blue and black are colours used in an earlier dictionary page. I have copied the page, but here (not in the glyph dictionary) changed 13? into *12 and adjusted for black *7 (etc) to see what happens:

 

The changes makes the table more complex, but the result is not impossible. Furthermore, new patterns emerge. We can, for example, add green 10 with red 10 to reach 20 glyphs which surround a center with 20 green and 8 red glyphs (= 28). Blue 10 will - inversely - be surrounded by blue 20 and black 20.

The use of colours in the G calendar will be no difficulty for a reader of the glyph dictionary, he will recognize the system from the K calendar.

In the G calendar we have no difficulty with the number of glyphs - all glyphs are intact. Our study of the K calendar has revealed, however, a possible complication in period 18 of the G calendar, 4 of the 7 glyphs may be black (or all 7 may be black). We should therefore investigate what possible effects that may have:

 

The change does not result in a convincing pattern. Twice 35 is 70 (and not 66). Counting with 7 black glyphs, on the other hand, will result in 35 + 28 + 7 = 70.

35 + 28 = 5*7 + 4*7 = 9*7 = 63.

We can still count 35 + 35 = 70 periods for periods 1-18, though. Introducing a black complication at the end does not necessarily change 70 into 63 (or 66).

In the K calendar we have here (not in the glyph dictionary) reduced the number of periods from 30 to 29, a change which seems reasonable due to the different type of end glyph in Kb4-18.

The 'black' glyphs are located immediately after viri in Kb1-14. The special type of end glyph exemplified in Kb4-18 is in reversed form located immediately before viri in Kb1-14. In Tahua we have seen the same kind of arrangement:

 

Long count gives 50 respectively 47 (not good), and common count gives 49 respectively 46. If counting should be done, it must be short count.

(With 13? instead of *12 as the number of glyphs in period 16 short count gives *46 instead of *45, a worse number because 2 * 45 = 90.)

We ought to look what effect a reduction in the number of periods - similar to that from 30 to 29 in K - may cause in G. Though in the dictionary page we are studying the opposite process is taking place - adding yet another period (33*).

I now notice how I have located the asterisk after the number instead of before, which is my convention for reconstructing numbers (as for instance in presenting number of glyphs per line in H).

Asterisk before a number means reconstructed, while asterisk after is used to make a comment.

The 'black complications' at the end of period 18 (in G) did not necessarily change the sum 70. If we regard period 32 as black, that does not necessarily change the beautiful structure presented in the dictionary page. Here (not in the glyph dictionary) I have separated the last triplet of periods into 3 lines to make the study more easy:

 

Number 3 (odd) in period 27 forces addition with period 28, resulting in 2+2+3+3 = 10. Number 3 (odd) in period 29 forces addition with the rest of the numbers, resulting in 3+4+6+4+3 = 20.

The inconvenience with black denoting not only the last two periods (32-33) but also the late autumn period (25-31) can be solved - if necessary - by citation marks: '32' respectively '33'.

The rhythmic 3 period groups presented in the G calendar was possible to use also in the K calendar, but the pattern breaks down in the 16th period if we change 13? to *12, because 55 ≠ 2 * 28:

 

Do I have to change in the dictionary because of this? No, I decide not to. It is enough with a question mark after 13 for the number of glyphs in period 16. The complications we have confronted as a result of changing 13? to *12 must, for pedagogical reasons, remain hidden until later on in the glyph dictionary.