diff --git a/README.md b/README.md
index 65b2036..1ab78a9 100644
--- a/README.md
+++ b/README.md
@@ -78,16 +78,6 @@ are trashed inside the block.
 Not because it saves 3 bytes, but because it's a neat trick.  Doing it optimally
 is probably NP-complete.  But doing it adequately is probably not that hard.
 
-> Every routine is falled through to by zero or more routines.
-> Don't consider the main routine.
-> For each routine α that is finally-falled through to by a set of routines R(α),
-> pick a movable routine β from R, move β in front of α, remove the `jmp` at the end of β and
-> mark β as unmovable.
-> Note this only works if β finally-falls through.  If there are multiple tail
-> positions, we can't eliminate all the `jmp`s.
-> Note that if β finally-falls through to α it can't finally-fall through to anything
-> else, so the sets R(α) should be disjoint for every α. (Right?)
-
 ### And at some point...
 
 *   `low` and `high` address operators - to turn `word` type into `byte`.
diff --git a/tests/SixtyPical Fallthru.md b/tests/SixtyPical Fallthru.md
index 11d701e..a27e998 100644
--- a/tests/SixtyPical Fallthru.md	
+++ b/tests/SixtyPical Fallthru.md	
@@ -5,6 +5,50 @@ This is a test suite, written in [Falderal][] format, for SixtyPical's
 ability to detect which routines make tail calls to other routines,
 and thus can be re-arranged to simply "fall through" to them.
 
+The theory is as follows.
+
+SixtyPical supports a `goto`, but it can only appear in tail position.
+If a routine r1 ends with a unique `goto` to a fixed routine r2 it is said
+to *potentially fall through* to r2.
+
+A *unique* `goto` means that there are not multiple different `goto`s in
+tail position (which can happen if, for example, an `if` is the last thing
+in a routine, and each branch of that `if` ends with a different `goto`.)
+
+A *fixed* routine means, a routine which is known at compile time, not a
+`goto` through a vector.
+
+Consider the set R of all routines in the program.
+
+Every routine r1 ∈ R either potentially falls through to a single routine
+r2 ∈ R (r2 ≠ r1) or it does not potentially fall through to any routine.
+We can say out(r1) = {r2} or out(r1) = ∅.
+
+Every routine r ∈ R in this set also has a set of zero or more
+routines from which it is potentially falled through to by.  Call this
+in(r).  It is the case that out(r1) = {r2} → r1 ∈ in(r2).
+
+We can trace out the connections by following the in- or our- sets of
+a given routine.  Because each routine potentially falls through to only
+a single routine, the structures we find will be tree-like, not DAG-like.
+
+But they do permit cycles.
+
+So, we first break those cycles.  We will be left with out() sets which
+are disjoint trees, i.e. if r1 ∈ in(r2), then r1 ∉ in(r3) for all r3 ≠ r2.
+
+We then follow an algorithm something like this.  Treat R as a mutable
+set and start with an empty list L.  Then,
+
+- Pick a routine r from R where out(r) = ∅.
+- Find the longest chain of routines r1,r2,...rn in R where out(r1) = {r2},
+  out(r2} = {r3}, ... out(rn-1) = {rn}, and rn = r.
+- Remove (r1,r2,...,rn) from R and append them to L in that order.
+  Mark (r1,r2,...rn-1) as "will have their final `goto` removed."
+- Repeat until R is empty.
+
+When times comes to generate code, generate it in the order given by L.
+
 [Falderal]:     http://catseye.tc/node/Falderal
 
     -> Functionality "Dump fallthru map of SixtyPical program" is implemented by