Re-Dating the New York Recording Laboratories' L- Series Matrixes
(Grafton, 1929 –1932):
An Update

By Alex van der Tuuk

This article is a follow-up on an earlier article on this subject, published in Vintage Jazz Mart #127 (Autumn 2002, pages 4–8). With new evidence coming in, I intend to update this article in order to pinpoint the dates of recording sessions held in Grafton, Wisconsin during 1929-1932. The original text was used for a presentation at the Canadian Collectors’ Congress in Toronto in April 2004.

It is possible that not all L-series recordings were produced at the Grafton facility. During a recent interview with Bud Shiffman (born 1912), he recalled that one the sessions he made with the Smyth-West Orchestra for the Broadway label was done in an abandoned ice house in Port Washington. This information throws new light on the subject and in the development of a timetable. If you have any additions to this article please contact me:


For more than a decade the New York Recording Laboratories (NYRL) used recording facilities in several cities, including New York, Chicago, and Richmond, Indiana. Besides recording their own artists in rented studios, NYRL leased or bought many masters from other companies to fill many series on the Paramount, Broadway, and Puritan labels.

By 1929 NYRL’s board of directors must have become aware how much cost saving a studio, close to their pressing facilities in Grafton, Wisconsin would offer. A nearly empty building across from the pressing plant, connected by a viaduct, was found suitable to house a recording studio on its second floor.

A problem in bringing African-American artists up to Grafton was finding locations where they could stay overnight in case sessions ran for multiple days. There was no large black community in either Grafton or Port Washington. There was no hotel for them, although reports have been made that some artists stayed at the Grafton Hotel. Artists were sent to Milwaukee’s black district, around Third Street, close to the railroad station, known as Brewers Hill. The Hill, at that time, was largely German, and in the 1920s the Third Street area was called “a veritable Baghdad in a mostly humdrum Milwaukee – Arabian night of adventure, and daytime bazaars of bargains.” Third Street was “The longest continious shopping street in America” and considered one of Milwaukee’s busiest commercial districts outside of downtown. Shops lined both sides of the street.

On King Solomon Hill’s recently found Pm 13125, “Times Has Done Got Hard,” Hill refers to this district:

From Brewers Hill, Milwaukee, I’m going back to Chicago.

Ishmon Bracey, together with Tommy Johnson, Charley Taylor, and Kid Ernest (according to Klaus Uwe Dür, this is probably Kid Ernest Moliere), stayed for about a week in a big boarding house on Fourth Street in Milwaukee during a recording session for the NYRL in Grafton. He referred to the area having taverns, open houses, different first-class places where they played.

The artists made their way to Grafton by interurban, and pressing foreman Alfred Schultz took them back to Milwaukee by car at the end of the day. African-American artists were not allowed to use the front stairs to get into the offices to transact business, and were forced to use the freight elevator to get to the second floor, where the office was.

It appears that black and white recording sessions were separated. None of the interviews I found refer to interracial recording sessions or even the presence of either black or white musicians at each others sessions.

For the recordings made at Grafton, a new master series was used: the so-called L-matrix series. The meaning of the L-prefix has been open to speculation for many years. One of the possible options was a reference to Arthur Laibly’s name, at that time recording manager for the NYRL, although he disputed this. One of Paramount’s 12000/13000 series talent scouts, Henry C. Speir from Jackson, Mississippi, thought the L-prefix referred to Lab. or Laboratories, an explanation that seems plausible since in 2000 I discovered a blank recording card at a local historical library. At the left corner, the information for the recording says “LAB. NO.”

Dating this series has been open to speculation for many decades due to the lack of surviving recording files. The standard discographies based their timetables on one specific date, supposedly written down on a test of Irene Scruggs, the notorious L-348-2 test. The date listed as May 28, 1930 has led to confusion, especially with the dating of surrounding recordings. According to an article in Matrix magazine from 1960, discographer Max Vreede reported that L-325 and L-326 were recorded May 28, 1930. James Asman had interviewed Scruggs in England in 1952, finding a double-sided test in her possesion of L-325, backed with “How I’m Feeling” (matrix 5127-2). It is likely that this test bore the date, as no test of L-348 was ever mentioned being in the hands of Scruggs.

No collector or researcher could tell me the whereabouts of test L-348. Things got even more complicated when Tom Freeland mentioned that a test pressing of Louise Johnson’s matrix L-398-2, apparently made at the same session, was in the possession of the late Irene Scruggs when she was living in Germany and bears the date 28 May 1930. Of course it is possible that the 4 was misread for a 9. But as late as 1980, no test bearing a date was found in the hands of Irene Scruggs, as reported by Dietrich van Staden. Scruggs died in Germany on July 20, 1981. No trace has been found of either her daughter, Leazar, or of the recordings.

The lack of this evidence makes it difficult to reconstruct a new timetable but appears not to be impossible.

The Start

For some reason the first recordings attributed to this series have an A-prefix instead of the L-prefix. It has been suggested that while the studio in Grafton was not yet operational, some recordings were made in Milwaukee by a freelance engineer. Although the Milwaukee city directories and Yellow Pages for 1929 to 1932 do not list any sort of recording studio, it is reported that one or two did exist in this city, although their locations are unknown.
Evidence for a non-Grafton recording may be the presence of the control numbers indented in the recording. The NYRL used control numbers for recordings not made by themselves.
The recordings A-1-2/A-2-1 by Papa Charlie Jackson (Pm 12853) were recorded in late 1929.
The record first appeared in the November 1929 Dealer’s List (which appeared late October) as “new release.” The first appearence in the Chicago Defender was November 9, 1929. Normally it would take four to six weeks between recording and release date.

The first L-prefix found so far is L-4, identifying W.C. Childers in Paramount’s and Broadway’s country or old-time series. Its first appearance was in the October 1929 Dealers’ List. Other early recordings in this new series bear control numbers: L-5, L-6, L-7, L-17 and L-18. This might indicate that the studio in Grafton was not ready for use. L-6/7 first appeared in the Chicago Defender of November 30, 1929 and as ''new release'' in the December Dealer' List. L-17/18 first appeared in the Chicago Defender of October 26, 1929 and as ''new release'' in the November Dealers' List.

Some recorded songs were copyrighted at the Library of Congress as late as April 9, 1930. It took some six weeks between recording and registering at the LoC. These included L-30, L-36, L-69, and L-75, as well as an unrecorded song (“Chain Store Blues”) registered under the name of W. R. Calaway, a talent scout. These recordings are listed in discographies as having been recorded circa September-October 1929. Most of the recordings between L-20 and L-100 were not issued earlier than April or May 1930. So what happened during the winter months of 1929-1930?

It is doubtful that winter conditions had anything to do with the delay in recording artists. Weather reports in the Port Washington Pilot for January to March 1930 do not indicate an extremely cold winter or extreme snowfall. There were some cold periods during these months, the last report dating from March 27 mentioning that country roads were blocked but traffic on the railroads continued pretty well on schedule.

It is more likely that the recording engineer at Grafton had difficulty in making the studio and its equipment ready to use, as H. C. Speir remembered being in Grafton in late 1929 to help reconfigure the studio. Janet Erickson also recalled seeing her father hanging drapes and towels in front of the walls to deaden the echo to improve recordings.

The Depression apparently hit the NYRL hard. While there was still a run on records from mid- to late 1929, by early 1930 sales were starting to decline. By March 18, 1930, recording artist Emry Arthur, who had started working for the pressing company in Grafton, wrote that hours had been cut down to eight hours a day and if the company did not get any jobs a lay-off was at hand. He alo made reference to songs being ready to record. Is it possible that Emry Arthur recorded his sides on L-105 to L-108 in March 1930?

Tommy Johson – Ishmon Bracey – Charley Taylor –
Kid Ernest

Based on interviews with Ishmon Bracey, conducted by David Evans, the session that comprised L-225 to L-254 is generally dated as circa December 1929. Bracey et al. stayed in Milwaukee for a week, including four days of recording. During the other days the studio was occupied. The evidence for dating this session lies in the fact that Bracey mentioned seeing a series of artists during a colored union gathering, including Blind Lemon Jefferson. Jefferson appears to have died on the night of December 18/19, 1929. So if Bracey saw Jefferson, they would have recorded before this date. Wardlow’s interviews with Bracey mentions the same artists, except he mentioned Blind Blake instead of Jefferson!

Bracey also mentioned that Louis Armstrong offered him a job as second guitarist in his orchestra after the death of one of his guitarists. Bracey felt unfit for the job even though he was offered six weeks to rehearse. David Evans mentions that this offer was made in 1930, according to Bracey.

A recent publication on the Jake Leg poisoning outbreak of late February 1930 by Dan Baum, Tommy Johnson’s “Alcohol and Jake Blues” (L-250), and Ishmon Bracey’s “Jake Liquor Blues” (L-225) all strongly suggest a March 1930 recording date. The illness was caused by consumption of Jamaica Ginger, a cheap medicinal liquor substitute that had been adulterated with an industrial plasticizing chemical for purposes of evading scrutiny by Prohibition enforcers. The poisoning of the supply did not even take place before late January, and then it was shipped out to various distributors. Jake Leg symptoms first appeared by mid-February and were first diagnosed in Oklahoma on February 27, 1930. Almost overnight many casualties were reported, with 1,000 victims in Mississippi alone. The poisoned ginger appeared to be highly toxic, especially to the spinal chord and peripheral nervous tissue, leading to temporary or permanent limber, or Jake Leg. The term first appears in print in a March 7, 1930, newspaper article in Oklahoma City.

Tommy Johnson mentions in his “Alcohol and Jake Blues” (Pm 12950):

Drinking so much of Jake, till it done give me the limber leg

Ishmon Bracey’s “Jake Liquor Blues” (Pm 12941) is even more specific:

If you don’t quit drinkin’ that poison Jake you’re drinking
It’s gonna leave you with the limber leg

Bracey’s “Jake Liquor Blues” was listed as “new release” in the June 1930 Dealers’ List, so the record was issued late May. It was Arthur Laibly’s policy to ask a recording artist if he knew of a song that would be a potential hit. He would do so with Son House (“Dry Spell Blues Parts 1 & 2,” Pm 12990) and Skip James (“Hard Time Killin’ Floor Blues,” Pm 13065). It is interesting that “Jake Liquor Blues” was Bracey’s first record issued. The outbreak of the Jake Walk would have been apparent, and it seems to have been a logical idea to issue this record first. Johnson’s record was issued a month later and appeared in the July 1930 Dealers’ List. Taking this into consideration, a session date of circa March or April 1930 is more appropriate than the traditionally cited December 1929 date.

The Columbus Prison Fire

John McGhee recorded this title in the Grafton studio. The song was based on a disastrous fire, which took place on Easter Sunday, April 21, 1930 in the Columbus, Ohio prison. The next day’s newspaper reported 319 deaths. The song, with matrix number L-290 may have been recorded shortly afterwards.

Henry Speir Reads from Puckett, Mississippi, Newspapers

Some important clues to dates can be inferred from two recordings made by Mississippi talent scout Henry Speir. According to Wardlow, Speir went to Grafton in late April or early May with the Delta Big Four, whose recordings can be found on L-312, L-313, L-315, L-316, L-318 to L-320, L-322, and L-340. These sessions embrace L-325 and L-326 by Irene Scruggs which were recorded around the notorious May 28, 1930 date.

An intervening matrix, L-314 (two distinct takes), is by Spier himeslf. On L-314-1 he reads from a newspaper dated April 19, 1930; on L-314-3 from a newspaper dated April 12, 1930

Dates Written on Test Pressings

Several tests found over the years by collectors sometimes bear written information on the white label, which appear to be processing dates, a quite common practice. Pressing foreman Alfred Schultz was one of the people involved in this. According to his daughter Janet Erickson, her dad worked on the white labeled test pressings in a rectangular, rather small room. It was located near the pressing room, possibly across from it. The room was a “filing” and “playback” room where the new records were previewed. Two phonographs were in the room, one at the right end of the room and one in the middle. The room had many little cubbyholes and shelves. They contained the manila file cards and ledgers on which Schultz would write.

Some test pressings give what appear to be recording dates.
2-3 A double-sided test pressing of Papa Charlie Jackson [L-25-2] backed with Blind Blake [L-23-3] bears this information. If it is indeed a date, it would equal a February 3,1930, date. If this assumption is taken for a fact, than this implies that Charley Patton and Henry Sims recorded in early 1930, as already stated by Gayle Dean Wardlow. Patton and Sims recorded in the spring of 1930, according to Sims' sisters. A February date would be in order.

T5/2 #1 When I’m Gone Don’t You Grieve – White male vocal with guitar
T 5/2 #111 Two unknown fiddle tunes – Violin and guitar
This appears to be a processing date for matrix L-273, “Don’t Grieve Over Me” by George Stevens, issued on Pm 3252 and Bwy 8268. If this is true, then the Johnson/Bracey/Kid Ernest/Taylor session could very well be dated late March or early April 1930.

A test of L-359 by Bud Spaight’s Harmony Kings exists with the following information written on the label: 6/14 # 2. Mark Berresford (U.K.) supplied an undated newsclipping stating that Spaight’s band made four recordings in Grafton “last Monday.” If 6/14 stands for a processing date of June 14 (1930), a Saturday, then the band’s session would likely have taken place on Monday June 9, 1930.

Another test of this group from a different session exists. It is a test bearing the marks T 7/10 # 1. The title is the same as that of L-372-1. This indicates an early July date for this session.

Son House’s test of “Walking Blues” has the information 9/2 # 2 written on the label, this indicating a late August 1930 session date, as was already strongly suggested by House’s “Dry Spell Blues,” which was based on the drought of that year. Son House mentions Blind Lemon Jefferson as a lodger during his stay in Grafton. Apparently he meant to refer to Blind Blake, who recorded immidiately after the Patton/House/Brown/Johnson session [L-437 to L-440].

L-540, a test of the Cardinal Trio, bears the date 10-24-30, equivalent for October 24, 1930.

Unknown matrices:
3/5 High School Marching Band
PM & WK 3-3 Pack Up Your Troubles/On Wisconsin – Paul Margret
PM 7-10 Old Black Joe – Paul Margret at the piano and Art Laibly at the recording machine

More Direct Evidence

For some recording sessions I have found dates that pinpoint several sessions. One of them is the Skip James metal mother which was in the John Steiner Collection for L-746-1. A date of 2/4/31 is etched in the rim, equalling the American notation of February 4, 1931. This brings the Skip James recording closer to the end of January 1931.

Other information comes from surving artists who recorded some sessions that were heretofore unknown. Most of them involve overlooked dance bands or polka bands from the Wisconsin area that are not everyone’s cup of tea. Information on two specific recording sessions was found in a 1931 notebook owned by Romy Gosz, later a renowned polka king. He made his first recordings in Grafton on April 20, 1931, and a second session on July 17, 1931. The matrix numbers belonging to these recordings show that recordings can be pre-dated by two months. The note book was on loan to Greg Leider from Fredonia, who reissued several LPs and later CDs of Gosz’ recorded output, including the 1931 recordings. The notebook has since been returned to Gosz’ daughter, who lives in Manitowoc.

Alfred Puls is another musician who provided a specific recording date for the band he played tuba in: the South Side Orchestra of Two Rivers. This small town is only 7 miles south of Manitowoc. He came up with a November 22, 1931 date for the recordings made at Grafton for matrix numbers L-1270 to L-1284. These recording were made shortly after the King Solomon Hill/Famous Blue Jay Singers of Birmingham/Marshall Owens/Ben Curry sessions and were close to Blind Blake’s forelast session. According to Wardlow, this is consistent with the theory that Blake returned to Florida for the winter months, returning in the spring to Chicago.

Two invitational letters from the NYRL to Sig Heller, a Milwaukee-based dance band, exist with titles requested for recordings and dated October 16, 1931 and June 16, 1932. Heller stated that the band recorded a week or two after receiving the first letter and that recordings had to be made at a Saturday or Sunday, as the boys were going to college on weekdays.

Bearing all this in mind, a new timetable can be drawn that gives us a better idea of the number of recordings that were made per month. These figures show that the NYRL recorded a fairly consistent number of masters per month, although in some cases more masters were made when artists from the South arrived in Grafton, as happened with Skip James and the Patton/House/Brown/Johnson sessions.


1. There was a slow start in September 1929 in the Grafton studio, possibly due to malfunction of the recording equipment.

2. A restart of recording activities was made in January 1930, around matrix L-23.

3. Gennett, Marsh, and Q-R-S masters were used for issues in the catalogs for November 1929 to April 1930. From then on, mostly Grafton recordings were issued.

4. Average production of masters in Grafton was 45 to 70 per month.

Alex van der Tuuk is the author of the highly acclaimed book Paramount's Rise and Fall, published by Mainspring Press. He has written for many jazz publications in the U.S. and Europe, and is currently completing a book on Midwestern jazz and dance bands that recorded for NYRL.

Site © 2009 by MAINSPRING PRESS, LLC. Article © 2005 by Alex van der Tuuk.
Additional content is copyright as noted. All worldwide rights are reserved.

No portion of the material on this site may be reproduced, altered, or distributed in any form or by any means without the prior written
consent of the copyright holder(s). Unauthorized use will be addressed under applicable U.S. law.

For permission to reproduce from any Mainspring Press online or print publication, e-mail the publisher.