That'd may be hard, because there's no canonical 'original' Lisp (there were a few similar, but incomplete/incompatible versions), and more importantly, there's no clear definition of what even is a Lisp.
I mean, from the ones shown, it's sort of obvious, but I'm thinking in general. Also, Common Lisp, Racket, and Clojure are different enough that they count as separate languages in their own right, rather than 'variations', which a dotted line might imply.
Logo is a Lisp and yet does not really use s-expressions, and even has infix operators.
Now, you might argue that means Logo is not a Lisp, but it is actually extremely similar to basic Scheme. You could turn an interpreter for one into an interpreter for the other with mostly minor tweaks. It's also a dialect of Lisp historically.
Qi uses S-Expressions but it's got strong typing, pattern matching, and optional lazy evaluation. Sounds a lot more like Haskell than Scheme to me.
I'm not sure there is such a thing as a "Lisp." If you mean S-expression language, just say s-expression language. Wedging in Lisp as a substitute conflates the issue.