The Atlanta records puzzle

e have to remember that in the discographies of 78 rpm discs the number or alphanumerical combination that chronologically identifies a recording is what is known as «matrix number». It is given in ascending numbers to the successive recordings of a given record company. Matrix is the physical element on which each particular piece was recorded.

As for the disc itself, there were three basic ways of singling it out:

a) The most common one was by means of the disc number. It consisted of a numerical value and the addition of a suffix for each side (generally A and B). For example, 180-A and 180-B. Each side was the result of pressing the respective matrix to which its number was given at the time of recording;
b) Also the face number was used. It was a different one for each side and even they may result not consecutive. For instance, one face might have been 65.135 and the other 65.246. Obviously, each face also represented, in this case, a different original matrix;
c) In some cases the concepts of matrix and face numbers were merged because each side had a different value but that coincided with the one of the matrix. For example, a determined disc may have on each side 60.123 and 60.145, both identical to the respective identifications of the matrices.

When around 30 years ago I began to analyze the particular case of the Atlanta records, at the beginning I took for granted the criterion followed by the researchers that had preceded me. They used to regard the number of each side not as a «face number» but as a «matrix number». We shall see it was a mistake.

There were some reasons that might have led to this error:
a) Those numbers were repeated three times on the disc and were easily seen printed on the record label and as an embossment in the area near the end of the recording. The third way in which they used to appear was hand-engraved on the original matrix in the area that was to be covered by the disc label and was noted as a bas-relief. As in other recording companies the matrix numbers appeared like that then maybe the belief that they also referred to them was reinforced in this case;
b) There was a listing used by the record collectors who had taken catalogues of that time into account. In it these numbers had been arranged in ascending order and correlative blocks corresponding to different artists were noticed. Something that was logical had they been, in fact, matrices.

The latter observation is based in the methodology by which the recordings are made. As placing the performers in front of the mouthpieces, waiting until they had tuned up or played something to «warm up» fingers or throat, etc. took some extra time, when they were ready to record takes they cut as many as possible one after another to avoid the waste of time. And, lastly, it meant money. Let us think if we alternatively made several performers enter and leave the recording studio, these idle periods of time with no benefit would be reiterated because the artists would «cool down» again, and we would have to put them in the right place and all else.



But all these considerations did not correspond with the numbers of the supposed matrices that were included in the final portion of the famous listing. In this portion there were no more than two or three successive values for each artist and there were many frequent cases in which to each value corresponded a different interpreter. I don’t know if other researchers were aware of this peculiarity and, if they did, no one mentioned it.

Then I began to wonder why to change the ways described above, transforming the recording studio into a kind of procession of players continuously coming and getting out of it, carrying their instruments, crowding at the door and waiting for their time uncomfortably in some little room nearby, to start again this absurd cycle. Furthermore, there were two series of such supposed matrices: the 65.000 with instrumental recordings (bands, orchestras, rondallas) and the 66.000 with voices (singers, poem readers, monologues).

They were two different artistic universes targeted, consequently, also to different buyers. By being matrices, their unquestionable characteristic of growing through time would indicate an initial saturation of the market with recordings of instrumentals (series 65.000), to later switch to voices (series 66.000). But in that way, during the first period, material may not have been offered to the fans of singing or those who liked monologues, consequently resulting in the loss of the buyers of that sector. Later they would have been fully satisfied while there was nothing new to be offered to those who wanted to go on buying instrumentals.

However, the advertisements at that time simultaneously announced discs of both types indicating a logical and foreseeable coexistence. This reality clashes with what was supposed in the preceding paragraph which, evidently, would have been an absurd case of marketing because no record market operated that way, not even in its beginnings.



An absurd situation we, however, shall finally reach if we consider those numbers as if they were matrices. Because of that the conclusion that something was wrong was derived. That approach «did not fit». We had not to follow what veteran record collectors and researches thought and then no longer regard them as matrices but consider those numbers as belonging to faces.

But how could we find out if there were other things that really represented matrices? This query led me to do something elementary that, because I fully trusted my reputable teachers, I had obviated by thinking the issue had been elucidated by them. So when I thoroughly checked the discs and their labels then a quite elusive alphanumerical series with a suffix z sprang up. (For example: 186 z). It was scarcely seen as a bas-relief in some of the labels and it seemed that nobody had taken them into account. Even though I succeeded in reading it in only a 40% of my records, when I arranged them in a correlative order the serials of faces 65.000 and 66.000 became interlinked. Now the true matrix blocks successively and alternatively recorded by orchestras and vocal interpreters were evidenced. With that, finally, a logical explanation to the simultaneous release of the two types of recorded material was found.

But there was still another doubt: why the initial 75% of the listing included face numbers that were, many times, consecutive for the same performer, or, at least, had numerical blocks for each one, in both series? And why towards the end of this first Atlanta period (when matrices were pressed in Germany) the listing seemed to be an apparent chaos?

A possible answer results by knowing how the commercial process was like. This record company label launched for sale a large amount of material in March 1913 with a wide repertoire in both series (which was the mentioned 75% of the total that was finally published, recorded restlessly from the mid- 1912. As the factory had at its disposal this huge amount of matrices by each artist before manufacturing the discs they had no trouble in numbering the faces with specific numerical blocks for the different performers.

It is not clear what was the aim of that but we know that by the late 1913 and in 1914, because the market had been saturated with the original catalogue, the company had need of new material. For that reason, possibly the final recordings by a wide variety of artists were sent in small lots. Because of that it was no longer possible to assign blocks to each one. Hence those gaps in the face numbers. Had they meant matrices they would have implied those above mentioned absurd times of coming in and out of the recording studio.

Confirming that the originally chosen system gave rise to confusions in its time we can mention that the subsequent releases had also a disc number in the disc labels. But that also would have reinforced the belief that the series numbers 65.000 and 66.000, still noticeable on the discs, meant matrices.

Would have they thought: Why not matrices if we have a disc number? But it happens at times that reality is more complex and deceiving than what we think. But that is an advantage for those who like to solve puzzles.