Appendix B: Columbia Matrix Series (1901-1934)

Columbia Master Book, Volume I, Tim Brooks, ed.

Foreign Language Recordings | Notes on Individual Series | Columbia Acoustic Matrix Series

Return to Appendices

Following is an overview of the matrix series used by Columbia during the acoustic era, with a few extensions into the 1930s.¹ Even though not all of these series are listed in the present volumes, it will serve to provide context and indicate the breadth of Columbia’s recording program. Readers seeking more information on foreign language series are referred to Richard K. Spottswood’s landmark seven-volume discography, Ethnic Music on Records. Note however that Spottswood lists only recordings made in the U.S.

These charts and notes will also be useful for approximate dating of odd-series Columbias, and for determining their probable country of origin. Unlike Victor, Columbia printed both catalog and matrix numbers on its labels throughout most of its history, so matrix numbers can readily be used for identification purposes.

No Columbia document has ever been found that explains the purpose of each matrix series. There must have been such a document, to keep the different recording programs from running into each other. Maybe it was a thick ledger stuffed in a safe behind Victor Emerson’s desk. Unfortunately, we have no alternative but to reconstruct the allocation scheme from actual examples found in different numerical ranges. This is bound to lead to some errors or mistaken assumptions, and additional information from readers is welcome.

Columbia’s matrix numbering will be examined up to the number 100,000—for all intents, the acoustic era (1901–1925). A few continuations are shown beyond that point, but I have not ventured into the bewildering jungle of prefixed series, which replaced the old matrix series in the 1930s. Several old series numbered below 100,000 were, in fact, reused during the 1930s with prefixes added. There was even a transitional period during which some masters received prefixes and others did not (e.g., in the 17,000s). All of this is beyond the scope of this essay. Also not included are a few irregular series of the acoustic era, for example the X-prefix (“Old Comrades March” on Columbia A1608, matrix X95) and S-prefix series (Ramon Garcia).

The size, origin and content designations given in the chart are the dominant ones so far identified for each series; there were of course exceptions. Where a series appears to have been limited to one or two languages, they are given. Otherwise the designation is “foreign,” with specific languages listed in the notes.

An upper limit for each block is given only if there are enough examples, or sufficiently scattered ones, to indicate that the entire range shown was used. Thus for the 7,000s no upper limit is guessed at, since no examples are known above 7049.

Foreign Language Recordings

Return to Appendices

Columbia was extremely active in recording foreign language material both for export and for sale to ethnic groups within the U.S., but it was not entirely consistent in assigning control numbers to these recordings. At the outset, in 1903, blocks were apparently allocated to specific foreign studios such as St. Petersburg (35,000s), Berlin (40,000s) and Paris (50,000s). Foreign language items recorded in the U.S. were assigned to the regular domestic series (1-4999). However this simple system soon broke down, and masters originating in the U.S. can be found in many foreign series along with those from abroad. For example, while some items in the Spanish 5,000s were recorded in Madrid or Mexico City, others were credited to “Señor Carlos A. Prince’s Banda Española”—obviously U.S. recordings by Charles A. Prince and the Columbia house band. Recordings by the U.S.-sounding “Columbia Orchestra” also turn up in foreign series, although this name might have been used as a cover for various nondescript foreign bands.

Matrix allocation of foreign material from around 1908 to the mid-teens is even more confusing (except for England, which always had clear cut matrix series of its own). All sorts of languages could be found lumped together in some blocks. For example, the 21,000s contain Spanish, Italian and Hawaiian material, the latter recorded in Hawaii! Foreign language recordings also continued to appear in the “domestic” series (19,100s, 38,100s) during this period.

By the mid-teens Columbia itself was evidently getting confused, so it lumped virtually all “foreign” masters into three series: London recordings, Spanish language recordings, and everything else. The general domestic series thereafter contained only domestic items, popular and classical.

It is possible that some blocks were divided into even smaller segments than are shown here. However the continual hop-scotching which that would imply makes little sense when continued for many years, and in many instances languages are known to have been thoroughly intermixed no matter how fine the blocks are cut.

Table 6a:

Columbia Acoustic Matrix Series

Return to Columbia Tables | Appendices

 * [Researcher Du Jun Min dates this marix block to 1905 and 1907 during two recording trips to China in an article in the APS Journal.]

** Based on evidence in Taylor, DAHR Editors believe this is multiple matrix blocks: an early block of 12" masters from ca. 1908-1914 assigned to Milan from 74700-74999 (with documented masters up to 74829); a 75500 block of uncertain origin ca. 1909; and three later 12" blocks, 74000, 75000, 76000, assigned to London (excluding the earlier numbers) dating from 1916 through 1923.

*** Evidence suggests that this is two matrix blocks, 36000-36205 St. Petersburg, 1907, numbers above 36250 are of Dutch origin, 1912-1914.

**** More detailed information on this block from Chris Zwarg:

35000-35188 St. Petersburg, 1st half 1903, session ending end of June
35225-35277 Warsaw, 2nd half 1903, probably continuation of above trip
35316-35478 St. Petersburg, reported as late 1903 but possibly later
35479-35690 St. Petersburg, 1906
35692-35999 gap filled with 1915 London sessions

***** 70000 is a London master but 70001-70999 are an Italian matrix block from 1920-1923. 

Table 6b:

Columbia Acoustic Matrix Series: Selected Series Above 100,000

Return to Columbia Tables | Appendices

Appendix B: Notes on Individual Series

Notes

1. This is a revised version of an article by the author that appeared in Record Research magazine in 1975.

Back to Appendix A: Matrix and Catalog Number Dating Charts | Forward to Appendix C: “Phantom” Matrix Numbers

Columbia Master Book Table of Contents | Columbia History | Tables | Appendices

The Columbia Master Book Discography, 4 Volumes, Complied by Brian Rust and Tim Brooks. Reprinted by permission.