I discovered an interesting relationship a few days ago (and I got a chance to be my own collaborator). I was generating rooted trees of nonempty sets with n points (i.e. rooted trees whose nodes are sets of 1 or more points.) So I worked out the first few terms 1,2,5,13,37,108,332 I deleted the initial "1" and sent it to the server and got: %I A029856 %S A029856 2,2,5,13,37,108,332,1042,3360,11019,36722,123875,422449,1453553, %T A029856 5040816,17599468,61814275,218252584,774226549,2758043727, %U A029856 9862357697,35387662266,127374191687,459783039109,1664042970924 %N A029856 Rooted trees with 2 types of leaves. %O A029856 1,1 %K A029856 nonn,easy,eigen %F A029856 Sequence is its own Euler transform. %H A029856 Transforms %A A029856 Christian Bower (bowerc@usa.net) I generated several more terms and saw that the 2 sequences were very likely the same (except for the first term). So I set out to find a correspondence between the 2 sequences. It turns out to be rather simple: To turn a rooted tree with 2 colored leaves into a rooted tree of nonempty sets, take the white leaves and join them into the parent node, so that a node with n white leaves as children becomes a set node with n+1 points. Similarly a node with is a set of at least 2 points can be turned into a node with n-1 white leaves as children. The leaves which are sets of 1 point become the black nodes. The bijection does not work for a one node rooted tree because it has a leaf which has no parent. This method also relates (non-rooted) trees with 2-colored nodes to trees of nonempty sets as long as they have at least 3 nodes (or points.) The tree sequences begin 1,2,3,7,14,35,85,231,... for trees of nonempty sets and 2,3,3,7,14,35,85,231,... for trees with 2-colored nodes. ___ | . | |___| __ _|_ | .| | ..| B W |__| |___| |/ _|_ _|_ B W W W . |...|| ..| \|/ |/ <---> |___||___| . . _\/_ \ / | .| R |__| Christian