September 3, 1999


By Tim Brooks

In the course of compiling volumes one and four of the Columbia Master Book Discography–the volumes covering the earliest, least-documented period of Columbia disc production–I was forced to confront a number of long-standing discographical controversies.[i]  Rather than simply repeat possibly erroneous information, each one was freshly researched, to the extent possible.  Whether or not you agree with the conclusions, the facts regarding some of these enduring mysteries are interesting.

#1: The Case of the Forgotten Brother

As most collectors know, Len Spencer (1867-1914) was one of the most prolific performers and producers in the early days of recording.  Beginning about 1890, virtually at the inception of the recording industry, he recorded first for Columbia, and then for the New Jersey Phonograph Company, Edison, Berliner, Victor and practically every other significant label until his death in 1914.  He had a deep, distinctive voice, and it is no wonder that when Columbia issued one of his specialties with a voice that sounded like his–perhaps even labeled as by “Mr. Spencer” or by “Spencer and Schweinfest.” Collectors assumed that it was Len.

However Len had a younger brother, Harry (born Henry) Spencer (1875-1946), who also recorded.  The number of records he made came as a surprise even to historian Jim Walsh, who, in his first, 1947 article on Len stated that Harry had apparently made only one solo recording, and assisted his brother on one other.[ii]  By 1958, when he published a follow-up article, Walsh had learned that Harry made many more than that, and could be found on at least five labels–Columbia, Edison, Berliner, Zonophone and Leeds.[iii]

In fact, Harry Spencer was a familiar voice on Columbia cylinders and discs during the late 1890s and early 1900s.  He was frequently heard bantering with Len on the latter’s minstrel records (as “Mr. Henry”), and from the late 1890s until 1902 or 1903 he appears to have been Columbia’s regular studio announcer, doing most of the spoken introductions.  He was also its principal artist for spoken word records.  Twenty-six sketches and monologues, on discs numbered between 19 and 880, are tentatively or positively identified as by him in the Columbia Master Book.  (There may well be others.)  The best known is probably “Address by the Late President McKinley at the Pan American Exposition” (833), which is identified as Harry in its Marconi issue.

Most troublesome, however, has been Columbia 21, the famous “Arkansaw Traveller” (sic), in which a country rube cracks jokes at the expense of a citified passerby while playing snatches of “The Arkansas Traveler” on the fiddle.  Numerous articles and liner notes attribute this best-selling title exclusively to Len, referring either specifically to Columbia or more generally to all principal versions.  The Edison and Victor versions are labeled as by Len Spencer.  The even more common Columbia version, however, is generally uncredited, or credited only to “Spencer and Schweinfest,” with no first names given.  The fiddler is clearly multi-talented instrumentalist George Schweinfest.  Most assume that the “Spencer” is Len.  But is it?

  1. As with the McKinley speech, the Marconi issue is specifically identified as by Harry Spencer.  His name is found both in the 1907 Marconi catalog and on the disc itself.  The Marconi release uses ten-inch Columbia matrix 21, take 12.  This is by far the most commonly found take, which is also found on most Columbia single-face and double-face (A406) issues, as well as on several client labels (Harmony, Lakeside, Oxford, Silvertone).
  2. Matrix lists compiled by Columbia librarian Helene Chmura in the 1950s, from now-lost company files, also identify the artist as Harry.
  3. Aural comparison also suggests it is Harry, not Len.  While the brothers had similar, deep baritone voices, Len had a florid, heavily modulated style of delivery, much like the mock tragedians he sometimes imitated on record.  Harry, by comparison, was usually pretty flat and emotionless, almost stiff. Play the Columbia version of “Traveler” side by side with the Victor or Edison and you’ll see what I mean.

Company files are now gone, so none of this proves conclusively that Columbia 21 is by Harry Spencer.  Some collectors will probably go to their graves insisting it is Len, because it “sounds like him.”  However I know of no copy that is labeled or announced as by Len, and the above evidence seems rather compelling.  Certainly the most common Columbia take was by Harry Spencer, and most likely, all of them were.

Although there are no recording files extant for pre-1910 Columbia, matrix 21 would presumably have originally been recorded in 1901 (it first appeared on Climax).  The only way to date the long-lived take 12 would be to determine the earliest label style on which that take is found.  This writer’s earliest copy of take 12 is on a disc bearing the black and silver label, with “Columbia Phonograph Co., Gen’l” arched across the top and a single prize shown next to the spindle hole–“Grand Prize Paris 1900.”  Sherman and Nauck, in their fine label guide Note the Notes,[iv] date this label (III.B.1) as in use only during 1905.  Does anyone have take 12 bearing an earlier label?

Other takes that have been reported (less frequently) are seven-inch takes 2 and 7, and ten-inch takes 1, 2, 3 and 13.

Ironically, it is possible–though not certain–that Harry Spencer, not Len, may have had the best-selling single version of this famous routine.  (Considering versions on all labels, Len would have sold more.)  Early Columbia pressings turn up often, and Columbia A406 remained in the catalog until 1926 (with artist credit given simply as ‘talking’).  The master was also issued on numerous other labels, which are frequently found.  Len’s Victor version was last remade in 1908, reissued on no other label, and I believe is found somewhat less often (although it too stayed in the catalog until the 1920s).  Comparative sales of Edison wax cylinders are difficult to judge, since they are fragile and not as many survive as discs.  Len’s four-minute Edison cylinder version was reissued on Blue Amberol 3745, but not on Diamond Discs.

The postscript on Harry Spencer is sad.  He seems to have done little recording after 1902, although he continued to work for his brother at the latter’s theatrical agency.  After Len died suddenly in 1914 Harry briefly ran the agency, but due to his mismanagement it dissolved.  He then started several unsuccessful businesses and worked for some years for the Pennsylvania Railroad.  Eventually he went insane and was committed to an asylum, where he wrote a raving letter denouncing his relatives and accusing them of conspiring against him.  He died there in 1946.[v]  Then, as now, he remained the “forgotten” brother, always in Len’s shadow.

#2: The Case of the Two Tenors: Ferruccio or Diego?

 Among the better known serious recordings on early Columbia are the 11 sides by Giannini, a mix of Neapolitan airs and operatic arias, two of them duets with Alberto De Bassini.  They were made in late 1903 and early 1904, and numbered between 1484 and 1771.  Catalogs and labels identified the singer only as “Signor Giannini,” but collectors and writers have always assumed that meant Ferruccio Giannini, the pioneering tenor who was the first “serious” operatic artist to make recordings, for Berliner in 1896, and later for Victor and Zonophone.

It was certainly a surprise, therefore, to discover that in later Columbia catalogs these 1903-1904 recordings were credited not to Ferruccio but to Diego Giannini, an obscure tenor then recording in the Columbia “E” series.  In the early 1900s Columbia did not routinely identify artists in its catalogs, and sometimes not on the labels either.  However with the advent of double-discs in 1908 this policy was gradually reversed, and in July 1912 the general catalog underwent a major overhaul, with many artist names added or expanded.  “Signor Giannini” became “Diego Giannini,” which it remained until the last of his discs was deleted in 1914.  Several collectors I contacted were surprised by the discovery, and suggested that since it was from an original source, it might well be true.

In this case, however, collectors are right and the Columbia catalog editors were wrong.  I subsequently obtained copies of some of the original 1904 single-face discs, and they are announced as by Ferruccio Giannini.  The announcements are delivered with a wonderful flourish (“Ferr-UUCH-io!”), in a heavily accented voice that is no doubt that of the tenor himself.  These announcements, which by the way are worth the price of the disc, had been physically deleted from the master by the time the double face versions were issued, and the 1912 Columbia editors probably assumed the artist was the only Giannini then recording for the company.  Diego, who was apparently not related to Ferruccio,  continued to record until the 1920s.[vi]

The moral?  Don’t believe everything you read, even if it comes from a period source.  If it seems illogical, look for additional evidence.

#3: The Case of the Mysterious Missus

 Cal Stewart was probably the best-selling comic monologist of the entire acoustic era.  His “Uncle Josh” yarns were the equivalent of a TV sitcom (Green Acres, perhaps), describing in loving detail the lives of the denizens of Punkin Centre–Jim Lawson the drunk, Ezra Hoskins the tavern owner, Si Pettingill the grocer, Hank Weaver, Deacon Witherspoon, and many others.  In 1905 the semi-serialized story took a new turn, as Josh took a shine to widder Aunt Nancy Smith, ultimately marrying her in “The Wedding of Uncle Josh and Aunt Nancy Smith” (Columbia 3058).  Aunt Nancy was also heard on other Uncle Josh recordings around this time, during their courtship and after the marriage, as in “Uncle Josh and Aunt Nancy Go to Housekeeping” (3229).

The labels and catalog listings for these early sketches credit them to “Mr. and Mrs. Cal Stewart.”  When they were remade in later years they were credited to Cal Stewart and Ada Jones, and many collectors insist that the woman’s voice was, from the beginning, that of Jones.  But why would Columbia disguise so famous an artist as Ada Jones, whose popularity could only have sold more records?  Once again we turn to historian Jim Walsh, who began writing about pioneer artists in the late 1920s and whose sources included people who knew and worked with Stewart, including orchestra leader Fred Hager.  In 1951 Walsh wrote,

“In the records made during the early 1900s the part of Aunt Nancy is taken by Mrs. Cal Stewart herself.  Mrs. Stewart was a Tipton, Indiana girl, Rossini Waugh, who is described by Fred Hager as a talented violinist.  She traveled with the troupe, which Stewart formed to give entertainments based on the Punkin Center characters.  Her brother, James W. Waugh, was also a member of the group, which Mr. Hager says was a great favorite in the Middle West.”[vii]

Walsh had evidently heard the rumors that the original Aunt Nancy was portrayed by Ada Jones.  Later in the same article he commented that “some collectors have the impression that records labeled as by ‘Mr. and Mrs. Cal Stewart’ were actually made by Stewart and Ada Jones, and even that they were married to each other, but this is wrong.”

Aural evidence confirms what Walsh told us nearly half a century ago.  Although Waugh and Jones had similar voice quality, the former’s voice is rather thin and harsh, and she giggled incessantly.  Jones sounded a lot more natural; Walsh observed that she developed an acidulous streak absent from Waugh’s more innocent conception of the role.  Playing a “Mr. and Mrs. Cal Stewart” disc next to one by Stewart and Jones quickly reveals the differences–to most collectors.

Waugh played the role of Aunt Nancy for several years.  She is not always billed on the label; for example her name does not appear on “Uncle Josh’s Courtship” (nos. 1906 and A396), although her voice is clearly heard.  The original Uncle Josh-Aunt Nancy productions are found only on Columbia, since at the time (1904-1905) Stewart was recording exclusively for that label’s cylinders and discs.  When the exclusive disc arrangement ended Stewart and his wife recorded the routines for Victor, in early 1907.

Most references are vague as to when Ada Jones took over the role of Aunt Nancy, implying that it was soon after this.  McNutt in his Cal Stewart biography says that it was in 1910.[viii]  However the earliest I can find Ada Jones billed on a record with Stewart is in 1919, the year he died.  Ada Jones seems to have taken the role only at the very end of Stewart’s career.  One of my favorites from this period, incidentally, is “Evening Time at Pumpkin Center” (A2789), in which the usually stuffy Henry Burr affects a ridiculous rural accent in a vain effort to fit in.

The real-life missus seems to have been Stewart’s preferred female partner on record.

#4: The Case of Madame Noe.

 The Sextette from Lucia di Lammermoor was one of the most famous and frequently performed arias of the early 1900s.  It was sung by all manner of operatic artists, burlesqued by vaudeville troupes, even arranged as a fox trot.  Irving Berlin’s comic version, as sung by Billy Murray with the Vaudeville Quartette on Victor 17119, is a delight (“That note alone is worth a dollar!”).  Or you could cut a rug with the cheery fox trot version by Paul Specht and His Orchestra on Columbia A3708.

Victor made a splash in 1908 with an all-star version, featuring major names from its Red Seal roster–Sembrich, Caruso, Scotti, Journet, Severina and Daddi (Victor 96200).  At the extraordinary price of $7 for one twelve-inch, single-faced disc it was the most expensive record in the Victor catalog, but it nevertheless sold well among status-conscious buyers, judging by the number of copies that survive.  Columbia, lacking Victor’s stellar talent lineup, responded as best it could with its own celebrity sextette recording in 1910, priced at $7.50 (Columbia A5177, later available as single-faced 30443 for $5.00).  The artists were Bronskaja, Freeman, Constantino, Blanchart, Mardonnes, and Cilla.  Victor later brought out new celebrity versions with varying lineups (but always including Caruso), in 1912 (96201) and 1917 (95212).

The plot thickened in 1920 when Columbia decided to remake its “celebrity sextette.”  The recording was released in January 1922 as single-faced 49768, and later on double-face 74000-D, 9014-M and other issues. This time the famous artists were Barrientos, Hackett, Stracciari, Mardones, Meader and Noe.

Noe?  While it is true that earlier celebrity versions usually included one or two voices of lesser rank (Severina, Cilla), Noe was totally unknown to collectors I surveyed.  The name seems to have been unknown to Columbia customers in 1920 as well.  It appears on no other known Columbia recordings, or for that matter on any label I have been able to trace.  Columbia said nothing about the singer in its announcement of the record.

The sextette was composed of a soprano (Barrientos), two tenors (Hackett, Meader), baritone (Stracciari), bass (Mardones) and mezzo-soprano.  By process of elimination, Noe must be the mezzo, so we know it’s “Madame.”  A first thought was that perhaps the catalog editors meant Doris Doe (I’m not making this up!), a rather obscure contralto who sang with the Columbia Light Opera Company later in the 1920s.1 But then opera expert Larry Holdridge came to the rescue.  He had wondered about the mysterious Mme. Noe as well, until turning up a clipping from the June 26, 1919, Musical Courier about one Emma Noe.  Larry wrote,

“She was with the Chicago Opera, singing contralto.  She then worked with Minnie Tracey who ‘mended her voice,’ and ‘made her over into a dramatic soprano.’  She was offered a Covent Garden engagement, but she couldn’t accept it as she was busy with other contracts accepted earlier.  Although she had been converted to a soprano, there’s no reason she couldn’t have done the recording, particularly as she probably knew the role.”2

Of course this does not prove that the voice on the Columbia recording was that of Emma Noe, but it seems likely.  After all, how many Noe’s do you know?

[i] Tim Brooks, The Columbia Master Book Discography, Volume 1; Volume 4 (with Brian Rust) (Greenwood Press, 1999).

[ii] Jim Walsh, “Len Spencer,” Hobbies, May 1947, p.18.

[iii] Jim Walsh, “Leonard Garfield Spencer as His Daughter Ethel Lovingly Recalls Him,” Hobbies, July 1958, p.30.  Walsh also remarked on his error in the February 1951, issue, p.22.  Harry Spencer’s Edison was “The Ravings of John McCullough”; the Berliner is unknown, but may be an untraced take of the same title on Berliner 633.

[iv] Michael W. Sherman and Kurt R. Nauck III, Note the Notes: An Illustrated History of the Columbia 78 rpm Record Label, 1901-1958 (Monarch Record Enterprises, 1998), pp.21-22.

[v] ibid., October 1958, p.31.

[vi] see Richard K. Spottswood, Ethnic Music on Records (University of Illinois Press, 1990).

[vii] Jim Walsh, “Cal Stewart,” Hobbies, February, 1951, p.24.

[viii] Randy McNutt, Cal Stewart, Your Uncle Josh (Fairfield, OH: Weathervane Books, 1981), pp.53,92.

1 Doe is listed in The Columbia Master Book as a member of the Columbia Light Opera Company ensemble on 50031-D and 50039-D, both recorded in 1927, and she could be on others.

2 E-mail to the author from Lawrence Holdridge, October 11, 1998.

This page was last modified on May 9th, 2016.
© 2011 Tim Brooks All rights reserved HomeTV HistoryRecord Industry HistoryCopyright IssuesConsulting ServicesBook and CD ReviewsAbout My BooksGeorge W. Johnson, the First Black Recording StarLinks & ResourcesDartmouth CollegePress RoomFAQ