%I #30 Nov 20 2021 21:27:18
%S 0,18,99,198,261,342,423,504,585,666,747,828,909,1089,1170,1251,1332,
%T 1413,1494,1575,1656,1737,1818,1881,1998,2772,2871,3663,3762,4554,
%U 4653,5445,5544,6336,6435,7227,7326,8118,8217,9009,9108,9999,10890,10989,11781
%N Let j = | i - i_written_backwards |, k = j + j_written_backwards; then k is in this sequence.
%C There were originally some terms missing from this sequence, for example 909 generated from 101001. It seems likely to me that Conway only checked generators up to 100000. It's possible to prove that there are now no missing terms of the sequence by noting that if an n-digit number is not a palindrome then the difference between it and its reversal is at least 9*10^(n/2 - 1). - _Oscar Cunningham_, Sep 10 2021
%D J. H. Conway, personal communication.
%H Oscar Cunningham, <a href="/A008920/b008920.txt">Table of n, a(n) for n = 1..30484</a>
%Y Cf. A004086, A335978.
%K nonn,base
%O 1,2
%A _N. J. A. Sloane_ and _J. H. Conway_
%E Offset 1 and missing terms added by _Oscar Cunningham_, Sep 10 2021