The 1922 HMV Pinafore
Chorus: Sydney Coltham, Harold Wilde,
Conductors: Harry Norris & G. W. Byng
Recorded in London,
For many years leading up to this recording, HMV had adamantly stuck to its policy of using its own house singers, rather than D'Oyly Carte artists, to record the G&S operas. Bowing to popular pressure, HMV finally agreed to put their toe in the water, so to speak, casting Derek Oldham in several tenor parts and Sydney Granville as Strephon in the 1922 Iolanthe. With Pinafore, HMV finally entered the water at least up to their waists. For the first time, D'Oyly Carte singers took a majority of the roles.
The Recording Sessions
The bulk of the recording was made during sessions on 31st May, 2nd June, 3rd July, and 27th July 1922. Two sides gave the Company considerable trouble, and an additional three sessions were held (29th October 1922, 16th February & 15th March 1923) to redo them. From the available facts, we may infer that finding an acceptable Ralph was the cause of some consternation.
HMV Album Cover
At the first session, on 31st May, Derek Oldham played Ralph. The only side he recorded that day was Side 3, containing his two solos early in Act I. It is difficult to imagine why he was replaced. Oldham made many distinguished recordings over a long career, and it is hard to believe that he was sacked after recording just two takes of one side. Perhaps Oldham withdrew for other reasons.
In any event, at the second session—just two days later—James Hay sang Ralph ("Can I survive this overbearing," Side 9). Hay recorded the remainder of the role on 3rd & 27th July.
Presumably, over the next three months, Rupert D'Oyly Carte and others reviewed the recording and found two sides unacceptable: Side 3 ("But tell me, who's the youth") and Side 14 ("Carefully on tiptoe stealing"). The one thing these sides have in common is the presence of Ralph, so one imagines that James Hay's performance was the cause of the problem.
On 29th October 1922, the Company assembled and re-recorded these two sides. James Hay sang Side 14 (which evidently was recorded first), but Walter Glynne replaced him on Side 3. Whether this was planned all along or decided in mid-session is unclear. Either way, neither take was succesful.
On 16th February 1923, the Company assembled again. Dewey Gibson, who otherwise made no G&S recordings for HMV, was now cast as Ralph. Nellie Walker replaced Bertha Lewis as Buttercup (presumably because Miss Lewis did not want to come in just to record three lines). Harry Norris, who had conducted the first five sessions, was also unavailable, and George W. Byng replaced him.
Dewey Gibson was not an acceptable Ralph either, leading to the final recording session a month later, on 15th March 1923. This time, Walter Glynne sang Ralph on both sides. Darrell Fancourt, who had been at the previous six sessions, was unavailable, and Frederic Hobbs sang his two lines in "Carefully on tiptoe stealing."
As was often the case with early recordings, roles were split among multiple artists. The multiple cast roles are allotted as follows:
- Sir Joseph:
- Fancourt sings "Never mind the why and wherefore"; Millidge sings the second-act finale. Fancourt doubles in "Never mind the why and wherefore."
- Glynne sings the opening aria and "Carefully on tiptoe stealing"; Hay sings the rest.
- Hobbs sings the two lines in "Carefully on tiptoe stealing"; Fancourt sings the rest.
- Jones sings the "This very night," "Carefully on tiptoe stealing" and "Farewell, my own"; Essex sings the rest.
- Walker sings only the recitative, "But tell me who's the youth"; Lewis sings the rest.
Ralph, Buttercup, and Deadeye probably would have been cast consistently, had it not been that Sides 3 & 14 were redone so many times.
The oddest bit of casting is Darrell Fancourt as Sir Joseph in "Never mind the why and wherefore." The record label credits only him, but HMV's artist sheets credit both Fancourt and Ranalow. The solo singer is certainly not Fancourt, but Chris Webster thinks he hears Fancourt's voice in the downward arpeggio in the last ensemble. It is also possible that this was an error, and Fancourt does not sing at all.
Victor Album Cover
George Low's Comments
George Low provided the following explanation for why Derek Oldham didn't, in the end, record Ralph:
On Wednesday, 31st May 1922, DOC were in Oxford, and after the recording session Oldham, Lewis, Granville, and Millidge (none of whom seem to have had a very heavy recording schedule that day), together with Norris, headed back to Oxford to perform in The Gondoliers. The next day saw a matinee of Yeomen and Ida in the evening, which may well have put a strain on Oldham's voice, because not only did he not travel up to London with the other singers for the second recording session on Friday morning, he did not appear in performance in Oxford that evening either (as Granville, Millidge, etc., did), understudy Charles Cecil going on as Frederic in Pirates. This suggests illness, perhaps vocal strain.
Perhaps Oldham was already feeling under the weather on Wednesday, his voice acceptable enough for a stage performance, but not to the powers that be at HMV. Perhaps he struggled through the two shows on Thursday, but had to admit double defeat on Friday, not being able either to record in London or to sing on stage in Oxford. (On Saturday DOC performed Iolanthe (matinee) and Mikado (evening), but the paysheets don't indicate who played Tolloller and Nanki-Poo: Oldham or Darnton.) (Maybe, of course, HMV had already decided to dispense with Oldham's services after hearing the results of Wednesday's recording session.)
Monday, 3rd July, was the first day of DOC's annual holiday and of course Granville, Lewis, et al, were available without any difficulties. But Oldham — even assuming that HMV were still interested in him — was a different matter. He had actually left DOC on Saturday, and in his memoirs he says he "went straight into Whirled into Happiness." Maybe his contract with the new management actually began on the Monday and he couldn't absent himself for a day's recording, or at least felt unable to do so.
Oddly enough, 29th October 1922 appears to have been a Sunday! The company were in Newcastle. Evidently Norris and the singers were able to travel down to London and back again.
On Friday, 16th February 1923, the company were in Birmingham, having performed Patience (matinee) and Ida the day before and about to perform Gondoliers that evening. Maybe DOC and Hay and Gibson and HMV reached an agreement as to who sang in the recording and who sang Marco on stage that evening.
On Thursday, 15th March 1923, the company were in Leeds. Ruddigore was played on both Wednesday and Thursday. Rollins & Witts only mark Darnton for Richard Dauntless this season, but I know that Hay had played the part the previous autumn. Anyway, neither Hay nor Gibson made the long journey to London and back. The only active DOC singer who did was Granville, who didn't sing in Ruddigore.
It seems to me that prosaic considerations of DOC performance schedules and railway timetables may well have dictated the casting at a number of early HMV recording sessions!
James Hay, who sings most of Ralph, had several stints with D'Oyly Carte between 1913 and 1926, including a brief guest stint as Ralph in January, 1922. As this was Hay's only D'Oyly Carte recording, it is appropriate here to insert some biographical details about him, kindly provided by Robert Morrison:
South Australian James Hay joined the D'Oyly Carte as principal tenor in May of 1914 and toured until September of 1914. A further stint with the company followed between July 1919 and June 1920, after which Hay returned to his homeland and signed on as principal tenor with the J. C. Williamson G&S Opera Co. for its Australasian tour of 1920-21. Back in Britain, Hay rejoined the D'Oyly Carte for its July 1922 - June 1923 tour, which proved to be his last with the principal repertory company, although he did tour briefly with D'OC's "small" company in July of 1925. Hay again returned to Australia for the J. C. W. G&S Opera Co's 1926-27 tour, during which time he directed the first production of Ruddigore to be staged by the company. (It received its belated Australian professional premiere at the Theatre Royal, Adelaide on Thursday, June 23rd, 1927. Hay played the role of Richard Dauntless and — as he was hardly likely to cut one of his own songs — the first Act duet "The Battle's Roar is Over" was retained and continued to feature in subsequent Australian revivals of Ruddigore. However, the other standard D'OC cuts and revisions to the score of the mid-1920s were followed.) During the company's Sydney season, the following biographical note was published on the theatre page of the weekly periodical The Bulletin, (February 3rd, 1927; page 36):—
James Hay, another of the Australians in the G&S company at Sydney, owns Clare (S[outh] A[ustralia]) as his birthplace; but he learnt to sing in the choir of the Anglican Cathedral at Perth [Western Australia]. After musical training in London he joined the D'Oyly Carte management in 1912. His first part was Colonel Fairfax in The Yeomen. So thorough was the drenching he had received in the G& operas that when he returned from Australia in 1921 he did Rackstraw in Pinafore at a moment's notice. He had called in at D'Oyly Carte's office merely to pay his respects, and while chatting with the entreprenuer word was received that Derek Oldham, who was playing the part, had fallen ill. Hay seemed to have been Heaven-sent; without him the show would have had to close down. In spite of his Australian breeding a slight Scotch burr clings to Hay's voice."
[Indeed a contemporary review of The Gondoliers staged during the season, published in The Bulletin, alluded to Hay singing "Take a pair of sparkling eyes" with a Scottish accent!]
Darrell Fancourt was the Company's principal bass for over thirty years, long enough to record most of the operas at least twice, and some of them three times. He joined the Company in 1920, replacing Frederick Hobbs, who evidently was recruited to sing chorus in several of the early recordings. It would thus have been logical for Hobbs to replace Fancourt in the one recording session he missed.
Bertha Lewis had a distinguished career in the contralto roles. She joined the Company in 1905, played the contralto parts for six months in 1909, then left until 1914. She then remained continuously until her tragic death from injuries suffered in an auto accident in 1931. She recorded most of her roles at least once (except for Dame Carruthers) and some of them twice.
Rollins & Witts say that Sydney Granville played Bill Bobstay as early as the 1908-9 repertory, under Gilbert's direction, but I am reliably informed by George Low that he played Deadeye (as an occasional substitute) in that Season, not Bobstay, although he did play it later. He played Captain Corcoran on stage with D'Oyly Carte on at least one occasion (18th February 1920, in Sheffield). He also played the role with the J. C. Williamson Company in Australia and New Zealand from 1926-28, after this recording was made. It has been suggested that Leo Sheffield, the stage Corcoran when this recording was made, was was afraid of the recording horn and bowed out at the last minute.
Violet Essex and Bessie Jones never appeared with the Company, but shared the soprano parts on many of the early recordings. Nellie Walker sang the soubrette parts on many of the same sets. Pamela Baselow and Edward Halland also never sang with D'Oyly Carte, but apparently were in the chorus for most of the early recordings and sang the occasional small parts.
Robert Morrison provided some background information on Frederick Ranalow:
Frederick Ranalow (1873-1953) was a popular operatic baritone who had first recorded for the Gramophone Company around 1910. He achieved great success playing the highwayman Captain Macheath in Nigel Playfair's long-running revival of John Gay's The Beggar's Opera, first staged at the Lyric Theatre, Hammersmith in 1920. H.M.V. recorded a selection of songs from the opera with Ranalow in December of 1920, which proved to be popular enough for the company to record a further selection in February of 1922, again with Ranalow as Macheath, and the Polly Peachum of the time, Kathlyn Hilliard, (George Baker's first wife.) (Sydney Granville too, appeared for a time in a later revival of Playfair's production of The Beggar's Opera when he took over the role of Lockit in February of 1928.) Ranalow also sang the roles of Ned Travers in recorded excerpts from Ethel Smyth's comic opera The Boatswain's Mate, (with former Savoyard Courtice Pounds as the Bo'sun), for H.M.V. in 1916 and Sharpless in the first English version of Madama Butterfly recorded by H.M.V. in 1924. (The latter recordings also featured Nellie Walker as Susuki, Bessie Jones as Kate Pinkerton and Edward Halland as The Bonze.)
In most of the early sets, the principal comic parts were taken by George Baker. Presumably he was unavailable for this Pinafore, accounting for the presence of Ranalow, who never appeared with the Company and had no background in these parts. Henry Lytton, the stage Sir Joseph of the time, may have been judged unsuitable for the recording horn, or perhaps he declined to participate for some reason. He would record the role in the 1930 Pinafore, however (with Granville, Fancourt and Lewis recording their roles for a second time).
Chris Webster compiled the following summary of cuts in the recording:
- Act One
- Overture — The roll at the beginning of the overture is cut. The overture starts from the last note of bar four.
- Opening Chorus — The orchestral reprise of the main theme is cut from the last sung note, so that this chorus is tagged off by just the last 2 and a half bars.
- No 4a — An extra chord is played after 'pleasing person' to round off side 4.
- No. 6 — "Over the bright blue sea." The last bar is cut so the segue into the Gent's chorus (no. 7) is from the penultimate bar.
- No. 7 — "Sir Joseph's barge is seen." The second half of bar 29, bar 30, and the first half of bar 31 are cut. This is at a side change and so one side ends immediately after the last sung note by the gent's chorus and the 'Gaily tripping' side commences from the second half of bar 31.
- No. 10 — "A British tar." The second verse is cut, so this chorus ends 'dead' after the first verse without the 'attitude' chorus.
- Finale Act One — The repeat of the orchestral play out at the end is not observed so this passage is played only once.
- Act Two
- Entr'acte — cut.
- No. 13 — "Fair moon." The first four bars are cut.
- No. 14 — "Things are seldom what they seem." The first 8 of the last 11 bars are cut, thus omitting Buttercup's 'dissemble/tremble' duetting with Corcoran.
- No. 16 — "Never mind the why and wherefore." The introduction into the 2nd and 3rd verses is cut from bars 3 to 6 inclusive. Also only the last 8 and a half bars of the play out are used because the first 8 or so bars of this play out are cut.
- No. 18 — "Carefully on tiptoe stealing." The first bar is cut and the second bar which commences this section appears to have a grace note (on 'G') added to the first chord (This sound horrible, but it is emulating the semi-quaver of the sung care-FUL-ly)
- No. 19 — "Farewell my own." The four bars marked 'brass' in the VS are omitted after the 'dungeon cell' chorus.
- No. 20 — "A many years ago." The orchestral link after the verse one chorus which leads into the second verse is cut.
- Finale — The chorus that concludes Buttercup's section is cut, thus giving a segue straight from Buttercup into Sir Joseph's section. The opera ends with just three chords (which I do not believe were written by Sullivan) after the final 'Englishman' chorus.
The CD Reissue
Sounds on CD and its energetic proprietor, Chris Webster, has re-issued this recording on CD, in a modern digital transfer that eliminates much of the original background hiss and brings these performances vividly to life. David Duffey posted this review:
In 1981, a colleague who knew of my interest in G&S sold to me for ukp5 the complete set of the HMV G&S acoustic albums. She thought she had duped me. No 'commercial' range of gramophones were then being sold capable of playing 78 rpm discs, and those machines which could produced sound just laughed at in those days of 'Hi-Fi'. On the cover of the topmost album in the pile I received was a gap-toothed sailor in a rope-bordered portrait frame. Judge then of my delight at seeing just the same face greeting me when I opened the packet containing latest offering from Chris Webster.
Having delighted in the rendering of H.M.S. Pinafore on my 78 sets, I was dumbfounded by the quality of reproduction achieved on the CD transfer. I have never, ever, been beguiled by quality of reproduction over historical/interpretive interest. Talk of woofers, tweeters and needle hiss always leaves me cold. In fact I prefer the sound of (say) Glenn Miller on 78 with a fibre needle to the same recording 'cleaned up'.
Chris's technical wizardry on the 1922/23 H.M.S. Pinafore, however engages my historical interest in D'Oyly Catre performance practice, because it removes extraneous noise and allows one to focus on the performance and interpretation. The issue comes complete with a critical commentary by Michael Walters, whose expertise is well known.
I recently prepared a tape for a friend performing Sir Joseph of no fewer than fourteen renditions of 'When I was a lad' Frederick Ranalow's was not one of them, but deserved to be. His reading of the role is 'untraditional' but valid. I disagree with Michael Walters's opinion that it is 'camp'; it ranks well above any of John Reed's recorded essays. It, of course, comes nowhere near Martyn Green's definitive interpretation.
The description of the recording of Ralph's role makes fascinating reading in the notes, and it is a fairly uninspired rendition we get — but then, Derek Oldam's voice would hardly have been any better that what we get here; similarly, the Josephine does not give an exceptionably distinguished performance. No, it is the D'Oyly Carte performers who engage most interest.
First and foremost the great Darrell Fancourt, just a year or two into the roles he would make his own and define. Interestingly sounding lighter voiced than both his later Deadeyes, and taking the upper alternative notes on 'ploughs the water' at a time when these were not published in the VS. On his next recording he takes the low G, but reverts to the higher on his last rendition. I wish I could remember which he did on the only occasion I saw him play the part.
It is interesting to hear the voice of his predecessor, Frederick Hobbs, singing, "Silent be". Fancourt's final recording of this phrase, and his "I told you so"'s, cannot be bettered, and it is interesting to hear how this interpretation develops over three recordings.
Sydney Granville doubles Corcoran and Boatswain. His Boatswain is of course familiar from the later Henry Lytton recording, although his voice is not prominent in the chorus in this as it is in the 'Lytton'. 'Englishman' can rarely have been song better in the 'traditional' style than it is here - although I have a sneaking liking for George Cook — and certainly the 1922/3 is better than Granville's later rendition. His Corcoran is, simply, excellent, although it is hard to credit that it is in fact Granny singing in the ensemble parts of Act II — and, in passing, impossible to imagine that it is Fancourt playing Sir J in the Bells trio as it is credited.
Bertha Lewis, and her deputy, Nellie Walker (not DOC) are exceptional as Buttercup.
This transcript is a must for anyone interested in the history of G&S performance practice. The quality of sound is exceptional, allowing the listener to appreciate the nuance brought to the roles by the classic performers in a way which hitherto was impossible for anyone who did not actually go to DOC performances in the inter-war years. The price of ukp15 compares well with current full-price CD issues, but, in terms of time, effort and expertise, certainly cannot recompense the producer.
|1978||Pearl||LP||GEMM 148/9||Includes 1927 Trial|
|1998||Sounds on CD||CD||VGS 202|
|1999||78s 2 CD||CD||GS09|
|1999||Opera Classics||CD-ROM||AE205||CD-ROM titled "From Which We Came," including numerous early recordings of opera. The G&S content (which is only a fraction of what's on the disc) includes the HMV acoustical recordings of The Mikado, H.M.S Pinafore, Princess Ida, and Ruddigore. Note that this CD can be played only on a computer, not on a conventional audio player.|
||31 May 22|
||31 May 22|
||15 Mar 23|
||31 May 22|
||31 May 22|
||2 Jun 22|
||2 Jun 22|
||27 Jul 22|
||2 Jun 22|
||3 Jul 22|
||3 Jul 22|
||27 Jul 22|
||2 Jun 22|
||15 Mar 23|
||3 Jul 22|
||27 Jul 22|
- Francis mentions only the release on HMV, not the release on Victor.
- Sources disagree on whether Harry Norris or George W. Byng conducted this set. Francis says it's Norris, but Chris Webster has determined that Byng conducted sides 3 and 14. Both of these sides required several extra sessions before acceptable takes were produced
- Sides 4 & 7 were included on Arabesque's LP reissue of the 1931 Ruddigore.
|Session One, 31 May 1922 — Hayes, Room 1. Conducted by Harry Norris|
|Issued s. 1||Cc1405-1
|Issued s. 5||Cc1406-1
|Issued s. 2||Cc1407-1
|Issued s. 4||Cc1409-1
|Session Two, 2 June 1922 — Hayes, Room 1. Conducted by Harry Norris|
|Issued s. 6||Cc1416-1
|Issued s. 7||Cc1417-1
|Issued s. 9||Cc1418-1
|Issued s. 13||Cc1419-1
|Session Three, 3 July 1922 — Hayes, Room 1. Conducted by Harry Norris|
|Issued s. 11||Cc1601-1
|Issued s. 10||Cc1602-1
|Issued s. 15||Cc1604-1
|Session Four, 27 July 1922 — Hayes, Room 1. Conducted by Harry Norris|
|Issued s. 8||Cc1745-1
|Issued s. 16||Cc1746-1
|Issued s. 12||Cc1747-1
|Session Five, 29 Oct 1922 — Hayes, Room 1. Conducted by Harry Norris|
|Session Six, 16 Feb 1923 — Hayes, Room 1. Conducted by George W. Byng|
|Session Seven, 15 March 1923 — Hayes, Room 1. Conducted by George W. Byng|
|Issued s. 3||Cc1408-9
|Issued s. 14||Cc1603-8
- The table above lists the order in which the sides were recorded. Where there were multiple attempts at a side on the same day, none of which are known to have been issued, the entire row is shown in italics, and the word "Rejected" appears in the "Result" column.
- Where there were multiple attempts at a side on the same day, at least one of which was issued, the word "Issued" and the side number appear in the "Result" column. For takes not known to have been issued, the matrix number is shown in italics.
- Details of unissued takes were supplied by Chris Webster, who had access to copies of the artist sheets made during the sessions. Where an artist is not listed on the sheets but clearly was singing, the name appears in [square brackets]. Where an artist may have been singing, the name appears in [square brackets] with a question mark.
- Webster has not seen the artist sheets for side 1 (recorded 31 May 1922). Two takes are known, since the second was issued. There could have been more.