
COMMENTS

Recursive transformation ENIPS for Catalan bijections has a welldefined inverse (see the definition & comments at A122204). For all Catalan bijections in A089840 that inverse produces a bijection which is itself in A089840. This sequence gives the indices to those positions where each ("primitive", nonrecursive bijection) of A089840(n) occurs "atavistically" amongst the more complex recursive bijections in A122204. I.e. A122204(a(n)) = A089840(n). Similarly, other "atavistic forms" resurface as: A122287(a(n)) = A122201(n), A122286(a(n)) = A122203(n) and A122202(a(n)) = A122284(n). See also comments at A153833.
There exists similar atavistic index sequences computed for FORK (A122201) and KROF (A122202). Both start as 0,1654720,... (see A129604). This implies that regardless of how complex recursive derivations from A089840 one forms by repeatedly applying SPINE, ENIPS, FORK and/or KROF in some order (finite number of times), all the original primitive nonrecursive elements of A089840 will eventually appear at some positions.
Other known terms: a(12)=65167, a(13)=65178, a(14)=65236, a(15)=169, a(16)=65302, a(22)a(44) = 1656351, 1656576, 1656777, 1656628, 1656704, 1659507, 1659538, 1659653, 1659798, 1659685, 1659830, 1660155, 1660582, 1660439, 1660476, 1660621, 1660196, 1661073, 1660930, 1660859, 1661004, 1661287, 1661360.
