A Skolem/Langford-like sequence with primes

 

September 18th, 2006

 

Hello SeqFan and Math-Fun,

 

Consider this sequence S:

 

S=2 3 5 2 11 3 2 17 5 2 23 7 2 41 5 2 11 13 2 7 5 2 47 29 2 17 5 2 11 19 2 13 5 2 23 31 2...

 

Definition:

 

Pick any term T in S (“11” for instance): there are “T” primes between this term T and the closest next T of S (left or right).

[There are indeed 11 primes between (any) two “11” in the sequence]:

 

S=2 3 5 2 11 3 2 17 5 2 23 7 2 41 5 2 11 13 2 7 5 2 47 29 2 17 5 2 11 19 2 13 5 2 23 31 2...

 

This Skolem-like prime sequence has been built according to those rules:

 

(I wanted only primes to appear in the sequence -- and possibly all of them)

 

Start with “2” and fill accordingly as many “boxes” as possible to the right:

 

S=2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . etc.

 

We *must* now find a place for at least two “3”; let’s try with the *first* empty “box”:

 

S=2 3 . 2 . 3 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 .  etc.

 

There is room for only two “3” -- a third “3” would have to be put in place of the underlined “2” -- which is not possible. If a third “3” is put somewhere else in an empty box to the right, we would have more than three primes between this third “3” and the second one. So we have only two “3” in the whole sequence.

 

We *must* now find a place for at least two “5”; let’s try with the *first* empty “box”:

 

S=2 3 5 2 . 3 2 . 5 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 . . 2 ., etc.

 

There is room for a pair odf “5” -- but can we add some more “5” to the right? -- Yes, plenty of them (infinitely many -- as we had with “2”):

 

S=2 3 5 2 . 3 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 ., etc.

 

We *must* now find a place for at least two “7”; let’s try with the *first* empty “box”:

 

S=2 3 5 2 7 3 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 ., etc.

 

It *doesn’t* work! There is no room for the second (at least) “7” (its box is taken by the underlined “2”)

Let’s forget this try and go to the *next* empty box with our first “7”. Is there room there for the second one? No, neither!

 

S=2 3 5 2 . 3 2 7 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 ., etc.

 

After another miss with the third empty box, we’ll have a hit on the fourth:

 

S=2 3 5 2 . 3 2 . 5 2 . 7 2 . 5 2 . . 2 7 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 ., etc.

 

Can we fill more boxes to the right with a few “7”? No -- so let’s proceed.

 

We *must* now find a place for at least two “11”; let’s try with the *first* empty “box”:

 

S=2 3 5 2 11 3 2 . 5 2 . 7 2 . 5 2 11 . 2 7 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 . . 2 . 5 2 ., etc.

 

It works -- and there is room for (infinitely) more “11” to the right:

 

S=2 3 5 2 11 3 2 . 5 2 . 7 2 . 5 2 11 . 2 7 5 2 . . 2 . 5 2 11 . 2 . 5 2 . . 2 . 5 2 11 . 2 . 5 2 . . 2 . 5 2 11 . 2 . 5 2 . . 2 . 5 2 11 . 2 . 5 2 . . 2 . 5 2 11 . 2 . 5 2 . . 2 . 5 2 11 . 2 . 5 2 . . 2 . 5 2 11 . 2 . 5 2 . . 2 . 5 2 11 . 2 . 5 2 . . 2 . 5 2 11 . 2 . 5 2 . . 2 . 5 2 11 . 2 . 5 2 . . 2 . 5 2 11 . 2 . 5 2 . . 2 . 5 2 11 . 2 . 5 2 . . 2 . 5 2 11 . 2 . 5 2 . , etc.

 

We *must* now find a place for at least two “13”; etc.

(etc.)

 

The sequence S will look like this after 2,3,5,7,11,13,17,19,23,29,31,37,41, have been placed (terms computed by Gilles Sadowski):

 

S=2 3 5 2 11 3 2 17 5 2 23 7 2 41 5 2 11 13 2 7 5 2 47 29 2 17 5 2 11 19 2 13 5 2 23 31 2 53 5 2 11 59 2 17 5 2 191 37 2 19 5 2 11 29 2 41 5 2 23 43 2 17 5 2 11 61 2 31 5 2 47 67 2 83 5 2 11 71 2 17 5 2 23 29 2 37 5 2 11 73 2 53 5 2 239 79 2 17 5 2 11 59 2 43 5 2 23 89 2 101 5 2 11 29 2 17 5 2 47 97 2 113 5 2 11 103 2 61 5 2 23 107 2 17 5 2 11 109 2 67 5 2 383 29 2 53 5 2 11 71 2 17 5 2 23 127 2 83 5 2 11 59 2 73 5 2 47 131 2 17 5 2 11 29 2 79 5 2 23 137 2 227 5 2 11 139 2 17 5 2 479 163 2 149 5 2 11 89 2 53 5 2 23 29 2 17 5 2 11 179 2 101 5 2 47 157 2 97 5 2 11 59 2 17 5 2 23 151 2 103 5 2 11 29 2 113 5 2 191 107 2 17 5 2 11 181 2 109 5 2 23 167 2 53 5 2 11 193 2 17 5 2 47 29 2 197 5 2 11 211 2 173 5 2 23 223 2 17 5 2 11 59 2 127 5 2 431 89 2 263 5 2 11 29 2 17 5 2 23 131 2 233 5 2 11 199 2 53 5 2 47 229 2 17 5 2 11 137 2 293 5 2 23 29 2 139 5 2 11 241 2 17 5 2 239 269 2 251 5 2 11 59 2 149 5 2 23 107 2 17 5 2 11 29 2 163 5 2 47 257 2 53 5 2 11 283 2 17 5 2 23 271 2 157 5 2 11 89 2 151 5 2 719 29 2 17 5 2 11 179 2 419 5 2 23 281 2 359 5 2 11 59 2 17 5 2 47 307 2 227 5 2 11 29 2 53 5 2 23 167 2 17 5 2 11 277 2 181 5 2 191 131 2 353 5 2 11 313 2 17 5 2 23 29 2 173 5 2 11 317 2 193 5 2 47 107 2 17 5 2 11 59 2 197 5 2 23 89 2 53 5 2 11 29 2 17 5 2 863 311 2 211 5 2 11 397 2 449 5 2 23 331 2 17 5 2 11 347 2 223 5 2 47 29 2 199 5 2 11 367 2 17 5 2 23 389 2 401 5 2 11 59 2 53 5 2 383 337 2 17 5 2 11 29 2 233 5 2 23 373 2 229 5 2 11 349 2 17 5 2 47 379 2 263 5 2 11 89 2 467 5 2 23 29 2 17 5 2 11 179 2 241 5 2 239 409 2 53 5 2 11 59 2 17 5 2 23 167 2 251 5 2 11 29 2 443 5 2 47 421 2 17 5 2 11 269 2 491 5 2 23 433 2 293 5 2 11 257 2 17 5 2 191 29 2 509 5 2 11 461 2 53 5 2 23 439 2 17 5 2 11 59 2 271 5 2 47 89 2 283 5 2 11 29 2 17 5 2 23 487 2 197 5 2 11 499 2 521 5 2 479 457 2 17 5 2 11 281 2 503 5 2 23 29 2 53 5 2 11 463 2 17 5 2 47 523 2 617 5 2 11 59 2 277 5 2 23 557 2 17 5 2 11 29 2 307 5 2 431 541 2 569 5 2 11 547 2 17 5 2 23 577 2 563 5 2 11 89 2 53 5 2 47 29 2 17 5 2 11 179 2 313 5 2 23 167 2 359 5 2 11 59 2 17 5 2  etc.

 

I’m not sure if there will be room for every prime; not sure either if the sequence is infinite (or correct so far)!

Is this of interest?

Best,

Ι.