Two-Part Code vs. One-Part Code: Degrees of Irregularity

I mentioned Karl de Leeuw's book the other day. I (re)read it because I remembered it described code in which figures runs vertically, while words are arranged horizontally. Such a two-dimensional arrangement introduces some irregularity without the need to have separate tables for encoding and decoding. I occasionally used the term "two-part code" to describe such a system, but I realized that when the code is "dense" (including many words with the same initial letter in one row), it does not seem appropriate to call it "two-part." I added a new sectioin on this terminology in "Code, Cipher, Nomenclator -- Notes on Terms in Cryptology".
I checked my usage of the term "two part" for this kind of arrangement. In most of the cases, I used the term with a reservation that the arrangement is not completely random. But I'm going to add a note about this in one of my previous posts: "Cardinal Mazarin Used Two-Part Code ".