An unexpected ingredient in the Cyclic Tag System was the C1-C1 pair, although one of the obvious extensions of the catalog of collisions which had been prepared was to move on to trinary and other multiple collisions. In the case of the C1-C1 pair, this should have been all the more urgent because it was a product of a reaction already in the catalog. The separation was too close to dismiss its collisions as composites of individual C-collisions.
On the other hand, full generality would have required studying collisions with all the other gliders; knowing a specific application, it suffices to study only EBar collisions, as shown in figure 8.
These collisions have to produce odd residues, thus at least one odd glider. The first is somewhat solitonoid since the EBar passes through leaving the C1's more separated than they were before. It is not evident how to go on increasing the separation, leading perhaps to a binary tree. The second is also solitonoid, passing the EBar and ejecting an A trimer, such as forms part of the (A-trimer,D1) leapfrog. It is also possible to get a single D1 glider, as the third collision shows.
The fourth collision is the important one, because it exploits the last EBar to close the cycle which began with the C3, and prepare the way for the next bit or the stop signal in the EBar stream.