[seqfan] Re: A121064

[Hans Havermann]

0, 0, 0, 4, 4, 6, 6, 3, 13, 22, 25, 15, 3, 12, 9, 18, 18, 9 (offset 0)

>> maybe the author only looked at values up to 100
>
> That was my first guess and the author's a(11) = 15 and a(12) = 3  
> are in line with that thought. Unfortunately, there are no a(13) and  
> beyond in that scheme.

I'll take this a step further. The a(10) = 25 would be correct for any  
counts up to and including the numbers 100 - 199. The number 200 adds  
another 10-letter instance to the count making it 26 at this point..

a(10) = 25 is correct only for counts up to and including 100- 199
a(11) = 15 is correct only for counts up to and including  98- 399
a(12) =  3 is correct only for counts up to and including  78- 299
a(13) = 12 is correct only for counts up to and including 610-2999
a(14) =  9 is correct only for counts up to and including 406- 409
a(15) = 18 is correct only for counts up to and including 405- 408
a(16) = 18 is correct only for counts up to and including 408- 439
a(17) =  9 is correct only for counts up to and including 308- 339

There is a modicum of consistency here: a(10) to a(12), a(14) to  
a(16); but the earlier consistency trumps the later one, so I think it  
fair to conclude that either a(13) to a(17) are superfluous to the  
author counting up to 100 (say) since a(12) makes *that* sequence full  
or, if it was the author's intention to go further, beyond 399 (say),  
then a(10) to a(12) are incorrect, in addition to some of the others.  
Take your pick.