ODJB - January, 1917 Columbia session?

Regarding the possibility that the Original Dixieland Jass Band recorded for Columbia on January 30, 1917, a few days ago I wrote about Mark Berresford's possition denying it, as was stated in his liner notes to Pioneer Recording Bands 1917-1920 (Retrieval CD, RTR 79043).

I am not sure when these supposed recordings were first mentioned, but Harry O. Brunn's book The Story Of The Original Dixieland Jazz Band (Louisiana State University Press, 1960) contains several paragraphs on this issue, starting with "less than a week after their spectacular opening at Reisenweber's (January 27, 1917) were under contract to make the world's first jazz phonograph record" and then describing how Columbia insisted that the band should record two popular songs of the time ("The Darktown Strutters' Ball" and "Indiana") instead of ODJB originals, and even how the small studio reverberated wall-to-wall with the sound, and "a gang of carpenters, who were building shelves in the studio, laughed and threw their tools about the room to contribute to the bedlam."

Checking Brian Rust's monumental discography Jazz And Ragtime Records (1897-1942) (Sixth Editon, CD-ROM version), you can check his comments on this ghost session:

"Many anecdotal sources over the years have cited January 24, 1917, as the date of a Columbia session by the ODJB, the results of which were supposedly resurrected for a belated September release on Columbia A-2297. However, no documentation of any such session has been found in Nick LaRocca's files or in the Columbia files, which logged the matrices used on that record on May 31, 1917".

I have recently discovered an old article by Brian Rust, that he first published in Needle Time #11 (July, 1987). In this article, he already brings what is, in my opinion, enough evidence to prove that the January, 1917 Columbia recordings did not exist:

-"reference to the original recording card in the CBS files for both titles reveals that four takes of "Darktown Strutters'" and three of "Indiana", two of each actually being used for issue to the public when the record (A-2297) was announced in the supplement for September, 1917. There is nothing on any of them to suggest that this was a test date, the products of which were dragged out of the "dead" files (Brunn, p. 71) and issued to counteract the success of a Victor date on February 26, 1917. When a band made a test, one take was usually enough; certainly not seven to cover two titles, and by an untried unknown quantity such as the Original Dixieland Jazz Band".

-"recently, I came into possession of microfilms of the Columbia artists' files, and these, arranged alphabetically, show all the titles, issued and rejected, by each artist from the Aarons Sisters to the Zoellner String Quartet. From the autumn of 1915 onwards, each title is noted with the date on which it was recorded, passed for issue (or rejected), date of issue (if any), and catalogue number (also). The sole entry under Original Dixieland Jass Band bears the date May 31, 1917 for both titles. Nowhere previously is there any reference to a test date, or any suggestion that the band had been in the Columbia studio prior to their Victor date. In the light of this discovery, it is evident that with their usual go-getter methods, Victor secured the services of the band for a test session on February 26, 1917, that it passed this test and the results issued with incredible speed under date of March 7".

On the light of recent research, now we know that Victor 18255 was not issued on March 7, but on April 15, but I think that, from both Brian Rust's and Mark Berresford's research, we can conclude that the Original Dixieland Jass Band did not record for Columbia on January, 1917.