Broadly this sounds like an indexed corporate diary of sorts, but his use of future reminders (or ticklers in the footnote) certainly points to the use of index cards in a Memindex-like fashion.

Keep in mind that he's writing in Britain and the Memindex from 1903 was a US-based product, though similar ideas may have been used at the time across the pond.

https://www.youtube.com/watch?v=vS-b_RUtL1A

LATCH method for search within note taking (Location, Alphabet, Time, Category, and Hierarchy)

that he used it little in later years. His author-title entries usually went into the main file, after the appropriate subject index cards.

This is a curious pattern and not often seen. Apparently it was Paxson's practice to place his author-title entries into his main file following the related subject index cards instead of in a completely separate bibliographical file. He did apparently have one comprised of six scantily filled drawers which he kept in his coat closet, but it was little used in his later years.

What benefits might this relay? It certainly more directly relates the sources closer in physical proximity within one's collection to the notes to which they relate. This might be of particular beneficial use in a topical system where all of one's notes relating to a particular subject are close physically rather than being linked or cross referenced as they were in Luhmann's example.

A particular color of cards may help in this regard to more easily find these sources.

Also keep in mind that Paxson's system was topical-chronological, so there may also be reasons for doing this that fit into his chronological scheme. Was he filing them in sections so that the publication dates of the sources fit into this scheme as well? This may take direct review to better known and understand his practice.

Frederic L. Paxson kept a collection of 3 x 5 " slips of thin paper that filled eighty wooden file drawers which he organized using topical-chronologic headings spanning 4639 BCE to 1949.

For creating narrative, Barbara Tuchman apparently relied on grouping her note cards by subject, person, or chronological sequence.

Lustig provides a description of some of the order of Gotthard Deutsch's zettelkasten. Most of it seemed to have been organized by chronological, geographical and thematic means, but often there was chaos. This could be indicative of many things including broad organization levels, but through active use, he may have sorted and resorted cards as needs required. Upon replacing cards he may not have defaulted to some specific order relying on the broad levels and knowing what state he had left things last. Though regular use, this wouldn't concern an individual the way it might concern outsiders who may not understand the basic orderings (as did Lustig) or be able to discern and find things as quickly as he may have been able to.

Goutor only mentions two potential organizational patterns for creating output with one's card index: either by chronological order or topical order. (p34) This might be typical for a historian who is likely to be more interested in chronologies and who would have likely noted down dates within their notes.