Completed East of England and The Midlands: Artists and Subjects, London: Artists and Subjects, Portraits: British 20th C 63 Can we resolve if this is a portrait of Lewis Evans by Philip Homan Miller or William Edwards Miller?

Topic: Artist

Is anyone able to verify if this painting of Lewis Evans is definitely by Philip Homan Miller? The History of Science Museum has a similar portrait painting, which is also on the Art UK which is by William Edwards Miller (c. 1852 - after 1929). We think it is highly unlikely the paintings were by two different artists. Any information about these artists and/or paintings would be extremely helpful.

Completed, Outcome

Thank you for contributing to this discussion, which is now closed. Unfortunately, from July 2024, Art Detective is being paused until further notice due to insufficient funding to continue running the service. All 887 discussions and more than 22,000 individual submissions remain accessible on the Art UK website, but no new comments can be accepted. This discussion may potentially be re-opened in due course.


The Collection has commented: 'We are unable to revisit the picture due to the present restrictions but I know we have previously looked at this discrepancy. The attribution to Philip Homan Miller is from the book, The Endless Web, written by Dame Joan Evans, Half-sister of Lewis. The book is the history of John Dickinson and his Company. The Oxford version appears to be a 'cut down' version as it contains less peripheral material than ours has.'

There seems little readily available information on either of the MIllers, and no obvious evidence they were related (which could also only have been as contemporaries given their dates).

P.H. Miller was born in Derry/ Londonderry, N.I. in 1845 and died in Marlow, Bucks., in 1928. The two works on Art UK other than this are a rather good workshop scene in a salt works c. 1870 and an indifferent mayoral portrait.

William Edwards Miller, born in 1851 (where?), was a prolific and polished society portraitist, including figures connected with both Oxford and Cambridge Universities, which would include Lewis Evans as a benefactor to the former. He was himself a noted antiquaran horologocal collector and museum benefactor who would certainly have known (or known of) Evans in that connection so there is no obvious reason to query the authorship of the Oxford version of the portrait - which was the family one.

That would leave two options: either that the attribution in Joan Evans's book of the current version (painted for the Dickinson Company) is a mistake and it is really a version by W.E.M. or it is a good copy made for it by P.H. Miller before the family one by W.E.M was cut down. 'Occam's razor' suggests the greater probability of the former, but where do we go from here?

It would probably help to have better information on both artists: there seems practically none online. P.H. Miller is recorded at '1 Campden Town Hill Road, Kensington' when he exhibited at the Royal Albert Hall with the London Salon of the Allied Artists Association in 1908 (5 works) and 1909 (2 works). W.E. MIller has clearly had some attention as a collector rather than an artist, but nothing more immediately obvious than this:

Just to add confusion there appears to be an American landscape painter also called William Edwards Miller (1872-1909).

Mark Wilson,

Actually I suspect the pictures really are by two different artists. If you look at the 30 artworks assigned to William Edwards Miller on ArtUK ( quite a few of the are explicitly copies of other artists (eg the portraits of Edward Stanley, Samuel Sanders, John Griffiths) while others are clearly posthumous and presumably based on old photos (E W Benson x 2, Sir Henry Lampton, Francis Douglas). So for him to copy another portrait would be normal.

Dame Joan Evans ( wasn't just Lewis Evans's half-sister (though admittedly 40 years younger) she was also a distinguished art historian and was writing the history of the family firm in 1955, with presumably full access to the archives. So she would be unlikely to get the name wrong of a painter. The only query over it being Philip Homan Miller would be why a not-very-prolific Irish painter was chosen for a boardroom portrait, though admittedly a slightly unusual one.

However a portrait of him from 1926:

suggests he was still active in the 1920s and Pieter's record of his death in Marlow, Bucks suggests his later years may have been near enough to Evans's base near Watford. He may have just been a very painstaking painter (presumably with other income) which would also explain why he wasn't asked to produce a copy for the family of what must have been a company-commissioned portrait, possibly to mark Lewis's retirement as Chairman in 1918.

For a copy, W E Miller would have been an obvious choice, especially given their shared interest in horology. It looks like the photograph presented in 1925 by Lewis Evans is of the copy rather than the Company's version which seems to be slightly more precisely painted though it's difficult to be sure. Comparison of the two paintings in real life would help.

The copy was inherited by Lewis Evans's brother Sir Arthur and the fact it was a copy may have meant that he felt fewer compunctions in cutting it down (though archaeologists may have their own opinions on Sir Arthur's practices about altering paintings). After his death in 1941, it only ended up with the History of Science Museum indirectly which suggests he was uninterested in the scientific content anyway. W E Miller seems to have been fairly punctilious in marking his copies as such and from where, but that information may have been lost when it was cut down.

Marcie Doran,

In support of Mark’s comment that the works are by two different artists, I am attaching a composite for ease of comparison.

I believe that I have found William Edwards Miller’s birth date, Pieter. There is a William E. Miller, “Artist (Portrait Painter) Rtd” living at 9 St. Petersburgh Place, Paddingon, London, England in the 1939 England and Wales Register on Ancestry. He is widowed, with a birth date of April 24, 1851. He is living with Emma Rothwell. She is a “Schoolmistress (Rtd)” with a birth date of March 17, 1876 and is single.

Jacinto Regalado,

So when did William Edwards Miller die? According to Marcie's findings, he was still alive in 1939.

Marcie Doran,

I suspect that he is the William E. Miller, age 88, born “abt 1852”, whose death in Marylebone, London, was recorded in the England & Wales, Civil Registration Death Index (on Ancestry) for January 1940 [February 1940, March 1940]. Deaths are registered quarterly.

Thanks to all for those addition: Mark's analysis accommodates the known facts rather better than mine, so until/unless provable otherwise it may be better to suggest (if two artists are involved) that PHM did the original and WEM the Oxford copy. It may be difficult to get further conclusively if it has to be done solely on stylistic comparisons given that work by PHM seems rather scarce. Perhaps the collection could, when possible, check for a signature or any other labels on the painting (or related company paperwork) if not already done.

Kieran Owens,

The Ulster Biography of Artists says of Miller: "Philip Miller was born in Derry and was educated at Queen's University, Belfast. He studied architecture and painting, and in 1879 began to exhibit at the Royal Academy, London. In 1890 he was elected to the Royal Hibernian Academy."

Week's News, of Saturday 16th December 1876, listed Philip Homan Miller as one of the prize-winners at the Royal Academy Schools, being rewarded with a premium of £10 for "drawing executed in the Life School during the year."

The Belfast Newsletter, of Saturday 22nd August 1891, carried the attached marriage notice, showing that Miller was married on the previous 15th August at Kensington to Marianne Sophia Holmes.

The attached extract from the 1901 UK Census records that Miller, aged 56, was living, as one of several residents, with "Marianne S(ophia)", 50, and one domestic servant, at 13, Pembridge Crescent, Kensington, in London. He is recorded as being "deaf". 13 is now a modern block, but 14 next door gives an idea of the style:

The attached extract from the 1911 UK Census records that Miller, aged 66, was living with Sophia, aged 55, childless, at 'Moyleen', West End Gardens, Marlow, in Buckinghamshire. The return also shows that the artist was "totally deaf".

The Dictionary of British & Irish Botanists records that Sophia Miller, "wife of Irish painter, Philip Homan Miller, exhibited flower paintings at the Royal Academy and at the New Watercolour Society."

Attached are other newspaper extracts that might help with a biographical profile.

Osmund Bullock,

Yes, Marcie has the right William Edwards Miller. Still of the same address in Paddington (but incorrectly recorded as 'William Edward Miller' in the probate index), he died 2 March 1940. His April 1851 birth was in Florence, Italy; he was the third son (of six, plus two daus) of Harriet (née Edwards) and John Miller, occasionally recorded as an artist himself, but basically a gentleman of private means, who seems to have been living on the Continent for a few years in the middle of the century when his second & third sons were born.

In 1871 William was a student at the Royal Academy Schools, as was his elder brother John Douglas Miller, later a well-known mezzotint engraver and protégé of George Richmond ( ), who was born in France in 1849 (the RA gives a slightly wrong birth year). His eldest brother Arthur William Kaye Miller (1848-1914) had just started working at the British Museum, of which he was later to become an influential Keeper of Printed Books ( ).

William’s profession is variously given as ‘Portrait Painter’ (1881), ‘Artist Painting’ (1891), ‘Picture painter’ (1901) & ‘Painting (Artist)’ (1911). He married in around 1880, his London-born wife Mary being in 1881 (but not subsequently) listed as a ‘Figure Painter’. I cannot at the moment find their marriage or confirmation of her maiden name. I’ll post all the documentary support (with its extra detail) shortly.

Kieran Owens,

Pieter, such a hearing disability would make for an interesting study. Other artists born deaf or becoming deaf during their working lives include:

• Joshua Reynolds (1723-1792) - His 1775 'Self-portrait as a Deaf Man' is at the Tate
• Francisco Goya (1746-1828) - Deaf from 1792 at the age of 46
• Thomas Arrowsmith (1771-1839), miniaturist
• James Martin (living in 1843)
• George Henry Wimpenny (1857-1939) - Hearing restored aged 75 after 66 years.
• Joseph Henry Sharp (1859-1953)
• Douglas Tilden (1860-1935) - American sculptor
• Alfred Reginald Thompson (1894-1979)
• Alexander McGregor (living in 1899) - Political cartoonist
• Hugh Grant (living in 1904)
• Martin Dutton (1921-2009) - aka 'Lizardman' - Woodcarver
• Chuck Baird (1947-2012)
• Harry R. Williams (1948-1991)
• Joseph Grigely (b. 1956)
• Muthukrishnan Ramalingam (b. 1957)
• Trevor Landell (b. 1960)
• Niall McCormack (b.1960)

No ladies in there, as none came up while looking, but there must be some.

Osmund Bullock,

Two minor corrections for the record: the birth in France of John Douglas Miller (William Edwards Miller’s elder brother) that I gave as in 1849 may well have been the very beginning of 1850, and it would require a birth certificate purchase to be sure which. Similarly his eldest brother Arthur was probably born in early 1849 rather than 1848 as I suggested. A certificate would also be needed to confirm that William married Mary J[ane] E[llen] Limbrick in 1880 (Q3 Pancras district), though it’s highly likely – despite her middle initials being given as ‘JR’ in the censuses after her marriage (and different again in her 1847Q1 birth registration). They had no children.

Attached are the supporting documents for the details already given, including evidence that William’s father John was also an artist, though in his first identifiable census appearance (1861) he is just a ‘fundholder’

Osmund Bullock,

Good find, but very puzzling. There is no England & Wales or Scotland marriage recorded for a William Miller between 1871 (when our WEM was single) & 1881 (when he was married) to any Mary with Jesse and/or Ruth (or initials) as a middle name. The 1880Q3 marriage to Mary Jane E C Limbrick in (St) Pancras RD fits in every other way. It was in the right area (censuses after their marriage give her birth place as Marylebone) and in the right year (the Apr 1911 Census says they had been married 30 years); moreover her birth year from the censuses and death record (slightly variable) is c.1845-48 (probably 1846-47), and there is a birth recorded for Mary Helena Limbrick at Marylebone in 1847Q1, and which probably ties in with an 1871 census record for a Mary Ellen Limbrick, born Marylebone 1846-47, living with her widowed father at Paddington. For non-Londoners I should explain that St Pancras, Marylebone & Paddington are/were adjacent areas/ancient parishes, and in the C19th their definitions for ecclesiastical, registration & census purposes varied, overlapped and often changed over time. See

In the absence of any other possibles, they look to me like they’re all the same woman, with a curiously flexible attitude to her middle names. Not that it matters much for present purposes, as all we’re really interested in is the basic biography of her husband.

Can anyone add to either man's exhibition record - i.e.venues and date-spans - , excluding Graves's RA listing to 1903/4 (SBA I can also check, but not immediately).

I find the RA's new website a practically unusable pain in terms of looking up 20th-century exhibition records, and other things are only on paper. PHM must have had a track record at the RHA in particular to become ARHA in 1890 and his wife Sophia is mentioned as having been reputed in Dublin as a flower painter, whwich suggests some sort of exhibition there before marriage.

PHM's striking ‘Workmen and Workwomen: interior of a Salt works’ ( presents a small puzzle of its own in that it was exhibited at the RA in 1885 and immediately presented to Salford Museum and Art Gallery 'in memory of Mrs Ellis Lever'. The why and how are unexplained, though presumably bought off the Academy wall for the purpose rather than presented by Miller. Perhaps Art UK could ask if they have any other information.

Marcie Doran,

I think the background of WEM’s wife Mary would have influenced his work. An Ancestry tree led me to the certificate showing her christening (attached). She was christened Mary Jesse Ruth Backhouse on February 4, 1846. Her parents were Henry and Margaret Backhouse. Henry was listed as an “artist”. In fact, both Henry and Margaret were artists.

There is a brief biography of Henry Fleetwood Backhouse (1818/1819-1901) here on the website “The Early Ruskin Manuscripts 1826–1842” and Art UK has one work by a Henry Backhouse, here:

A lengthy biography of Margaret Backhouse (née Holden)(1818-1888) is on Wikipedia, here: She was “a successful British portrait and genre painter during the 19th century”. The entry states that “Backhouse exhibited at the Royal Academy between 1846 and 1882. Between 1848 and 1885, some 80 works by Backhouse featured in Society of Women Artists exhibitions and she also showed thirty works at the Royal Society of British Artists in the same period.”

S. Elin Jones,

Attached is an obituary from the The Londonderry Sentinel - 29 December, 1928. It gives a very brief biography of the life of Philip Homan Miller, his life and marriage. However it also mentions that he was educated in Queen’s College, Belfast and the Royal College of Surgeons, Dublin, before changing his desired path from medicine to the Arts (architecture and painting) It also mentions that Sophie was “a noted flower painter”

As far as I can tell, Sophie may have exhibited 7 watercolour paintings over 5 Royal Academy Exhibitions and Philip exhibited 9 paintings over 8 Royal Academy Exhibitions.

Philip Homan Miller
1883 - Miller, P. H. 1 Colville Mansions, Bayswater, W.
428. Mrs Miller - Philip H. Miller

1892 - Miller, P. H. 8 Gloucester Mansions, Harrington Gardens, S. W.
1572. Fons Bandusae - Philip H. Miller

1895 - Miller, P. H. 13A Sumner Place, South Kensington, S. W.
1328. Maud - a portrait. - Philip H. Miller

1896 - Miller, P. H. 13A Sumner Place, South Kensington, S. W.
1499. “She that won’t play must be made play” - Philip H. Miller

1898 - Miller, P. H. 13A Sumner Place, South Kensington, S. W.
1527. “At the forest’s verge” - Philip H. Miller

1899 - Miller, P. H. Pembridge Crescent, W
1476. “Can it be true?” - Philip H. Miller

1903 - Miller, P. H. 1 Campden Hill Rd, Kensington W.
1390. Mariana in the Mounted Grange - Philip H. Miller

1915 - Miller, P. H. ,A.R.H.A 1 Campden Hill Rd, Kensington W.
537. Waiting - Philip H. Miller A. R. H. A
826. Verge of the forest “How full of briers is the working-day world” (Shakespere) Philip H. Miller A. R. H. A

Sophie Miller
1892 -Miller, S(Mrs), Gloucester Mansions, Harrington Gardens
1075. Pansies - Sophia Miller

1893-Miller, S(Mrs), 13A Sumner Place, Onslow Crescent S.W.
1027. Primulas - Sophia Miller

1894 - Miller, S(Mrs) 13A Sumner Place, Onslow Crescent S.W.
994. Peonies - Sophia Miller
1028. Christmas Roses - Sophia Miller
1171. Cluster Roses - Sophia Miller

1915 - Miller, S(Mrs), 1 Campden Hill Rd, Kensington W.
424. Magnolia - Sophia Miller

1916 - Miller, S(Mrs), 1 Campden Hill Rd, Kensington W.
1068. Peonies (a flower show) - Sophia Miller

1 attachment
Osmund Bullock,

Well done again, Marcie; my Mary Limbrick theory was a complete red herring. I saw two Ancestry trees giving Mary Backhouse as William Edwards Miller's wife, but because neither of them gave primary evidence of when and where they married, and I could find none myself, I unwisely chose to ignore it (Ancestry trees can often be completely wrong). However your further details led me to the ODNB entry for Margaret Backhouse which also gives the marriage, so I persevered...and now I know why the event was so elusive.

William Edwards Miller married Mary Jessie Ruth Backhouse on 20th May 1880 at the Vice-Consulate in Boulogne-sur-Mer, and afterwards at the Anglican Church of the Holy Trinity in the same place. The church's records before 1906 seem to be lost, but the marriage is shown in the overseas consular marriages index (attached). The exact date comes from a contemporary report in the Evening Standard (also attached).

One further tiny snippet about WEM is that he was an FSA (Fellow of the Soc. of Antiquaries).

Has anyone got easy personal access to Anne Stewart's 'Royal Hibernian Academy of Arts : index of exhibitors, 1826-1979' (1987) for the tally of P.H. (and possibly Sophia) Miller appearances there?

Pieter, I have now got Vol II of the RHA Exhibitors book in front of me. The brief details are:

Exh 1880 to 1928, from addresses 1880-81, 55 Torrington Square, London WC; 1882-86, 1 Colville Mansion, Powis Terrace, Bayswater; 1887-88, 124 Stephen's Green (Dublin I believe but it doesn't say so); 1889-91, 20 Lincoln Place (Dublin, no detail); 1892 at 8 Gloucester Mansion, Harrington Gardens, London SW; 1893-98, 13A, summer Place, S Kensington; 1899-1901, 13 Pembridge Crescent, London W; 1902-28, 1 Campden Hill Road, Kensington.

During this period PHM exhibited a total of 98 works at the RHA. I will mention a few of them: 1885 'Mrs Miller'; 1904 'Mrs P H Miller'; 1910 'Sophia and P. H. Miller'; 1925 'Mrs Miller'.

Turning now to Mrs Miller she exhibited at the RHA 1892-1921, a total of 29 works. Her addresses mirrored those of her husband during that period.

If more detailed information is needed please let me know.

Marcie Doran,

Pieter, regarding your questions (31/08/2021 @ 10:59) about PHM’s “Workmen and Workwomen: Interior of a Salt Works” of 1885 (, I believe that I have found the correct Ellis family.

According to a book that references an article in the Manchester Evening News (May 11, 1911), “Ellis Lever (1833-1911) was the youngest son of a miner-farmer from Kearsley, Lancashire. After selling wire ropes to collieries and manufacturing brattice cloth, he became a wealthy supplier of coal and coke for gas works, developing outlets throughout Britain and overseas. During the cotton famine, he helped organise relief committees and sewing schools for women.” (See John, Angela V. “By the Sweat of Their Brow: Women workers at Victorian Coal Mines”. 1980.

He had a key role in a gas contracts scandal that originatedin Salford in 1889. (See Moore, Dr. James. “The Transformation of Urban Liberalism: Party Politics and Urban Governance in Late Nineteenth Century England”. 2006.

According to his 1885 publication that includes his exchange of letters with the Home Secretary, Ellis Lever wished to reduce preventable accidents in mines and provided funds in 1883 (and offered funds in 1885) to support that work. (See Macdonald, Alexander. “Death in the Mines; Explosions in Mines”. 1885.

The Lever family, whichincluded seven children, lived for a time at Culcheth Hall, Bowdon, Cheshire. The matriarch of the family, Catherine Henshaw Lever (née Orrell) (born abt. 1834 in Ashton-under-Lyne, Lancashire), passed away on October 29, 1885. I have attached her probate record from 1889. Ellis Lever married Ada Mary Wormald de Burgh-Lawson in 1889 and he passed away in Conway, Caernarvonshire, Wales on May 2, 1911.

Ellis Lever was often mentioned in the “Manchester Courier and Lancashire General Advertiser”. His wife’s death was announced in that newspaper on October 30, 1885. I found them by searching for “Culcheth Hall”.

Thanks to both.

The RHA information suggests Miller was living in Dublin c. 1886-91: that is, in the slot he was not showing at the RA, nor apparently featuring in the England (spring) census of 1891. It may be where he met (or re-met) his Irish-born wife even though they married in London in August 1891.

While Catherine Henshaw Lever (née Orrell, b. c.1834 - d. 29 October 1885) looks very likely to be the 'Mrs Ellis Lever' in whose memory Salford received Miller's 'salt-works' painting, she died late in the year so one can't assume the picture came straight off the RA summer exhibition wall expressly for that purpose (though the acquisition date says it was within 1885). Only Salford is likely to be able to explain the 'how', but it's not really relevant here.

Martin Hopkinson,

The Royal Society of Portrait Painters and its contemporary rival's exhibitions should be checked

Here's another P. H. Miller puzzle. His -and his wife's - exhibiting address at the RHA (and elsewhere) from 1902 to 1928 (and to 1921 in her case) was 1 Campden Hill Road, Kensington. That was apparently just inside the Campden Hill junction with Kensington High Street, nearly opposite the tube station. The sites on both sides of the Hill junction are now covered by later mansion blocks, and even if no 1 was not demolished for them the surviving earlier terraced houses continuing up the Hill on both sides are very big.

At the same time we know the Millers' home from at least 1911 - plus resident cook, ladies' maid and housemaid in that year's census - was in Marlow, Bucks., not Kensington

It doesn't sound a credibly 'town and country house' situation - at least for a man who (even in 1928) only left something under £1600 at probate, so there must be some other explanation: i.e. who was actually living there from at least 1911 on?

Marcie Doran,

I think this March 2004 report “1 Campden Hill, Kensington. Architectural & Historical Assessment” by Alan Baxter & Associates provides the answer (see pages 13, 19 and 23 of 74 on the pdf). The house was built “shortly before the First World War” for Colonel Edmond Hills (later Grove-Hills)(1864-1922) and his wife Juliet. He was “a keen astronomer and president of the Royal Astronomical Society from 1913 to 1915”. She remained in the house after his death until 1927.
I’m not sure about the period before WW1.

Thanks Marcie, but that appears to be 1 Campden Hill, rather than 1 Campden Hill Road which runs down to meet Kensington High Street with the low numbers at that (south) end. Apologies if my use of 'Hill' rather than 'Road' helped confuse the issue.

Marcie Doran,

No apology needed, Pieter. It was my mistake - London street names are confusing.

According to the attached article from the May 23, 1919, edition of “The Vote”, the house “Moyleen”, in Marlow, was the “country home” of P.H. Miller and his wife.

1 attachment

Thanks again: I suppose one can't argue with that assuming 'studio' also meant London residence. There is an earlier element of doubt on that in the 1893 report of their studio 'At Homes' which is the last attachment in Kierans post of 29/08/2021 10:39.

It identifies their 'charming studio' where these took place as then in South Kensington, but implies their residence was separate in 'Onslow Crescent': that however is also wrong (it doesn't exist) since they then lived at 13A Sumner Place albeit that is very close to Onslow Square and Gardens etc - but all also in South Ken.

Based on present street numbering it looks like nos 1, 3 and 5 Campden Hill Road were demolished to build the 1930s mansion block called Phillimore Court that now fronts Kensington High Street, with Phillimore Walk behind it. The next house into CH Road south of that today is no 7.

Whichever way you look at it no. 1 must have been a large establishment in a prime location, which makes it at least misleading that their 1911 census return (plus staff) was for Marlow - though presumably because where they were on the night. We've not yet seen one for 1921, or street directories or electoral rolls that might clarify further.

In January 2010, David Johnson, Curator of Horology at the British Museum, gave a talk to the Antiquarian Horological Society on W.E, Miller. The announcement for it includes a portrait of him by Melville Fisher RA.

The subsequent Society members' report on it (attached) kindly provided by my former horological coleague Jonathan Betts includes useful information on him, not least that the Emma Rothwell with whom he was living at his death in 1940 was an 'adopted daughter'. I have yet to check if she was doing so at any previous census dates.

Could the Museum of History of Science be a bit clearer on the provenance steps for their version of Lewis Evans?

The accession data says 'Lent by Arthur Evans per the Department of Western Art, Ashmolean Museum in 1984' with a reference number of 1984-12.

Given that Lewis Evans's brother, the archaeologist Sir Arthur Evans, died in 1941, it would help to be clearer on who the 'lending' Arthur is/was: it clearly wasn't Sir Arthur in 1984, so -if from him- when was it originally lent to the Ashmolean, and is it still a family loan?

The likelihood of it being a family copy after the (probably PHM) original of 1925 is of course greater if it was done after Lewis Evans's death in 1930: before that WEM could have painted him from life, not least since they would certainly have been known to each other as collectors.

The attribution of a late-life portrait of W.E. Miller shown two comments above (04/09/2021 10:12) to 'Melville Fisher RA' seems suspect: I can find no such artist, only the American landscape painter Hugo Melville Fisher (1878-1946).

A first shot at potting PH Miller is attached: W.E. Miller can wait until there are more views on attribution of the intact version of the Evans portrait, as above.

The simplest explanation is that his half-sister Dame Joan Evans was simply wrong in attributing it to PHM, but so far the only fairly obvious support for that is the muted palette (shared with his 'salt-works' picture at Salford): WEM's -including of the cut-down version- is generally more warmly colourful. On-screen colour and handling comparisons based on 'imaging' are of course problematic: one really needs the two canvases side-by-side.

If she was right, however, then Mark Wilson (28/08/2021 14:06) has provided a scenario that seems to fit what facts we so far know.

1 attachment
Martin Hopkinson,

Melville Fisher - perhaps really Melton Fisher [1859-1939]?

Mark Wilson,

The Department of Western Art at the Ashmolean have Sir Arthur Evans's archives:

So the link was there. The 1988 note on the picture ( says "The Evans family have recently deposited the portrait on loan to the University of Oxford, and it hangs above the sundial collections in the Museum of the History of Science". Possibly this was done via the Ashmolean which may have taken other loan objects relating to their interests who then passed the portrait on the the History of Science Museum. This must have been in 1984 (hence the accession number) and so was fairly recent and likely to be correct.

Though it's possible that the phrase "Lent by Arthur Evans per the Department of Western Art, Ashmolean Museum in 1984" may not refer to Sir Arthur at all, as Lewis had five children by his second wife ( and this could refer to one of his sons or grandsons named after their eminent uncle. Though you would expect them to have gone directly to the Museum that Lewis helped found rather than via the Ashmolean (maybe they were too embarrassed at cutting down the picture).

The photograph of the unvandalised picture presented by Lewis in 1925 has a display label attributing it to W E Miller (, which suggests the copy was made by then - indeed it's possible the W E copy was made shortly after the boardroom one to be presented to Lewis. If Lewis had commissioned a separate portrait, it would have been as easy to get W E Miller to produce a new one.

The History of Science Museum, Oxford have commented:

'The Portrait of Lewis Evans, by W. E. Miller, c.1920 has been on loan to the History of Science Museum as stated on our database since 1984 from the Ashmolean Museum. However, the painting was gifted to the Museum in 2020 and at that point a considerable amount of research was undertaken concerning provenance and this is when the version by Philip Homan Miller at Frogmore Paper Mill & Visitor Centre came to light.

The portrait of Lewis Evans at the History of Science Museum was transported by Mr Arthur L. Evans Esq. (1930-1995) in 1984 from South Africa to the Ashmolean Museum. The portrait was one of two loaned at this time; namely the portraits of Sir John Evans and of Lewis Evans. The latter was transferred within the University of Oxford in 1985 to the then Museum of the History of Science as being more appropriate to its collection. It has remained on display in the History of Science Museum to date. Lewis Evans was the grandfather of Arthur L. Evans Esq.. The ownership of the painting passed to Arthur L Evans second wife Ursula Evans (nee Brown) in 1995.

The History of Science Museum archive holds a photograph Inv.13359 of the painting in an earlier state, before it was cut down and the background overpainted The details of when and why the painting was altered are still unknown. This photograph was presented by Lewis Evans to the Museum in 1925.

We have liaised with Frogmore Paper Mill re their painting but unfortunately COVID prevented any further research or visits.'

Osmund Bullock,

Google Books has a snippet view of Joan Evans' 1955 book 'The Endless Web': This suggests there is only one mention of Philip Homan Miller (or similar, or of the portrait) in it, and that is in the list of illustrations at the beginning of the book ("30. Lewis Evans. From a painting by Philip Homan Miller, 1910"). See attached 1. The illustration itself is not viewable, but I think the date is new to us. However on Ancestry there is attached to a pedigree of the Evans family a photograph of the portrait that is clearly the illustration from the book, as it is titled '30. Lewis Evans'. Attached 2.

So the first point to be made is that this means the information about the original portrait's painter did not necessarily come from the author herself. Had she written about the portrait in the text I would agree with Mark that she'd be unlikely to have got it wrong; but it is quite usual for the editorial team to deal with much of the legwork involved with indexes, lists of contents and illustrations - the last may need to be not just sourced, but also copyright-cleared. She could perhaps have known it was by 'Miller' (people of her generation commonly referred to artists by their surname alone, and C20th art was not her area of interest at all), but may have left it to the publishers to find out the details. It would still be a bit surprising if they got it wrong, but not out of the question - bear in mind that by 1955 the sitter and both Miller artists were long dead, PHM over 25 years before. And remember that the *only* source (so far) for the PHM attribution is the book.

Osmund Bullock,

Now that we have a full group of images to play with it is worth making some detailed comparisons. I usually make side-by-side double image for this, but in this case the two versions are so close that we need to look at pairs or trios of images and flip rapidly between them for differences. So I've prepared some, all cropped as accurately as I can to show the same view, and each should be looked at as a separate comparison exercise. Forgive me for the length of what follows, but it's hard to explain in any other way.

Osmund Bullock,

The first to be made is between the History of Science Museum's reduced version (which I will call 'HSM oil') and a cropped image of our portrait at Frogmore Paper Mill ('FPM oil'). This is useful because it will enable us to spot clear differences which may also appear in the black and white photos - and understanding what those photos show is what I'm trying to do. Don't worry about slightly-shifting angles & shapes, thinner faces, etc - photographs often distort such things, and as we've previously found, images on Art UK are often mysteriously (and sometimes majorly) distorted in uploading. The quick-flip method reveals many changes, but a lot of these will be less clear or differently-lit in the B&W photos, and some may have been affected by the overpainting and possible retouchings. See attachments 1 & 3, which you'll need to save so they appear the same size in your viewer.

I would ask you to focus on just three areas: (a) the top of the instrument he's holding relative to his waistcoat (and the gold watch chain just showing), (b) the position of the third waistcoat button (counting from the top), and (c) the right-hand end of the handkerchief showing in his top pocket. All three of these change significantly as you flip between them.

2 attachments
Osmund Bullock,

Next I will add the B&W photograph given to the History of Science Museum by the sitter in 1925 ('HSM photo'), again cropped to the same point - attachment 2. I suggest you save the B&W photo to a position in the middle of the two previous images (1 & 3) and flip between all three. Concentrating on the same three areas, to my eye it is clear that the B&W photo corresponds not to the FPM oil (our 'copy'), but to the HSM oil - it is, as the HSM themselves have stated, a photograph of *their* portrait before it was cut down and overpainted.

Actually you can do this without downloading - they appear the same size in the tabs in my browser, but flipping between them quickly is harder to do.

1 attachment
Osmund Bullock,

Finally do a quick-flip comparison between the photo from the (Endless Web) book illustration ('EW photo') and the 'HSM photo' - attachments 4 & 5, with slight crops to match. To my eye they show the same painting - indeed they are probably the same photo, though the book illustration quality is poor.

2 attachments
Osmund Bullock,

So if I'm right this leaves us with a problem: the cut-down family portrait and the photo of it when intact that are held by the History of Science Museum, and believed by them to be by William Edwards Miller - presumably on the basis of information from the sitter and/or his later family - is the same portrait as that illustrated in the 1955 book by Joan Evans and stated there to be by Philip Homan Miller. If PHM was involved at all (which I don't rule out, though the shared surname is worrying), then I can only think he painted the boardroom painting at Frogmore, and information from their archives supplied the name which was wrongly (but understandably) applied to the photo of the other one used in the book. Or perhaps their (our) version is signed: could we please see the highest-res image available to Art UK of the murky area bottom right?

Thank you Osmund, that is an incredible piece of work.

One thing that strikes me is the almost total absence of work by Philip Homan Miller on the internet. Even Artprice has nothing. Artnet shows one work but records that the artist was born in 1928!

As far as I am able I have reviewed the respective exhibition records of the two Millers. It appears that WEM is almost entirely known for his portraits, many of them no doubt will have been commissions. PHM had a much broader range of work, of which by my reckoning about 40% was portraiture (some of which were family portraits).

WEM was known as a respected painter and as far as I am aware wasn't a copyist of other artists' works. In regard to the copying of the portrait of Lewis Evans I suspect that he was commissioned to paint a copy of his original painting. The History of Science Museum considers it unlikely that two different hands are involved. As has been noted previously, the confusion may have arisen due to a possible error in the book by Dame Joan Evans.

What appear to be new facts are

(1) the 'Oxford' canvas is of direct descent from the sitter to his grandson Arthur, who lent it in 1984, the loan being converted to a gift from his family in 2020: Sir Arthur Evans, the sitter's older brother, has no part in the matter and should be written out of the current Art UK 'information'.

(2) The date given for the painting reproduced in Dame Joan Evans's 1955 book is 1910, but only as caption data - as is the artist attribution there to P.H Miller. There is no textual discussion of it so neither necessarily originated from Dame Joan's 'family knowledge'.

Osmund suggests that the book illustration may in fact be from the 'Oxford' canvas before it was cut down, not the one under discussion, and we know such a (negative replicable) photo already existed by 1925.

We have no independent date for the Frogmore/Dickinson version captioned as by P.H. Holman in the 1955 Dickinson history: all we otherwise know is that the sitter became Chairman of the company until 1918 (when he was 65) - which would have been the obvious date to obtain a 'boardroom' retirement portrait of him.

There is inherent credibility in believing W.E. Miller as artist of the Oxford canvas: even setting aside appearances he and Evans were both notable horological collectors who would certainly have known each other that way. One has to assume that Evans himself commissioned -or at least agreed the commissioning - of the Oxford 'family' version and, be it copy or original, WEM is the obvious person he would have favoured.

If it was/is in fact the original and done in 1910 all one has to accept is that it shows Evans at 55 rather than 65, and that it was regrettably later cut down and adjusted round the edges - probably for practical reasons that could have included shipping at some unknown point to South Africa, but not likely to have been until after his death in 1930.

The possibility all that suggests is that it is the Frogmore/Dickinson version which is the copy. There is no other reason that the Dickinson company should have commissioned one full of instruments, even knowing of Evans's passion for them, unless for presentation to him. Neither appearances nor probability suggest it ordered two at the same time (c.1918), one for him and one to keep!

If the Oxford one was indeed painted in 1910 for Evans, however, it would probably have been long known of by senior Dickinson people and Evans may well have given the company a print or prints of the same photo that he gave to the University with his collection in 1925: the negative for that could of course have been shot at any point after 1910.

By the time of the 1955 company history the Oxford one -in whatever state - was probably out of sight in South Africa and B&W photo turned up in the Dickinson archives, perhaps with a 1910 date on the back, could easily have been mistaken as being of the company version.

The issue is then how PHM's name comes in, since Evans could easily have sent back his original (and/or photos) to WEM for a copy to be made in about 1918 - as to a friend/ fellow clock collector as much as because he was the original artist. Despite the Oxford one being a different tone there are examples in the 30 by WEM on Art UK which are similar to that of the Dickinson version.

Unless there is clear documentary evidence (signature or paperwork), it begins to look as if the PHM scenario is more likely a mistake. PHM died in 1928, two years before Evans, having last exhibited at the RA in 1915 but continuing at the RHA to 1928. WEM had faded into retirement for at least ten years before his death in 1940. His current 'after 1929' death date on Art UK suggests that is the last known dated work (though among the Art UK tally it seems to be 1922). The proximity of PHM and Evans's deaths (and perhaps public notices of) may have led to such an error - easily perpetuated in technical book production as Osmund has already noted.

(In the recent discussion of Charles Frederick Tomkins (1798-1844) he turned up similarly misidentified in a standard modern biography of the actor W.C. Macready. He appears there only by surname but wrongly indexed as 'P.W. Tomkins' (i.e. Peltro William, 1759-1840), probably just because the dates fitted and absence of better information.)

I suppose we have to wait and see if the collection has anything to add but I agree it starts to look more 'mistake' than otherwise.

Martin Hopkinson,

The Courtauld Institute Library has copies of the catalogues of many of the annual exhibitions of the Royal Society of Portrait Painters which might be relevant, and of some of the exhibitions of the Modern Society of Portrait Painters of the second decade of the last century
It is quite likely that the Royal Academy Library and the National Portrait Gallery hold runs of these catalogues too
When the National Art Library reopens of course one could check all of them

WEM's only RA exhibit after 1903 was a 'presentation portrait' of the Scottish Civil Servant, Sir William Baillie Hamilton, in 1909.

The other possibility not yet mentioned (I think) is that if WEM did the original of and for Evans before 1918, the latter may have had him make the Dickinson copy on his retirement as Chairman, either at the board's request (which would presume board cost) or his own to present to the company. The PHM attribution and lack of record as such a gift suggest the latter is perhaps unlikely.

Osmund Bullock,

And according to Johnson & Greutzner (British Artists 1880-1940), at *none* of the venues listed by Grant did WEM's exhibiting extend beyond 1909.

Mark Wilson,

Regarding WEM not being a copyist, in my comment on 28 August, I pointed to three of WEM's entries on ArtUK where he was producing a copy of somebody else's work. There's copies of George Richmond, Lockhart Bogle, G F Watts and possibly Joseph Jacobs, mostly explicitly marked as such and done for Oxbridge Colleges. There's others where it looks as if he was copying off photographs and the dates make it impossible it was from life (at least one, of R E Lawley, is marked as such).

It surprised me as well, given that he seems to have had no problem getting direct commissions, but maybe there was less stigma about doing such work at the time for an established artist. Him producing a copy, either of his own work or someone else's would be pretty standard. And although he may not have exhibited much after 1909, there are several painting dated after that on ArtUK, though possibly from earlier photographs.

Osmund's detailed comparison confirms my hunch that the photo given in 1925 was of the WEM family one prior to its being cut down. It's what you would expect as being more accessible to the family. The pre-existence of a negative might have made that the easiest option to illustrate the 1955 book as well. As it was a fairly faithful copy, I don't think that Dame Joan would have been too worried about it not being the exact original when the original creator was credited. The book was about the firm rather than the art after all.

The question is how much reliance can be put on the picture caption in that book. You'd expect a professional art historian who authored a lot of books to be more accurate than average, even without the family connection and pick up an editorial mistake if it was someone else's work. And it would be an odd mistake to make to pick a fairly obscure artist instead of a more prolific one. In addition W E Miller seems to have signed his work fairly prominently (including copies) and doesn't normally do such elaborate backgrounds - normally just some wooden panelling, though his own tastes might make this an exception.

It will probably need examination of the Company records and the Dickinson Boardroom picture to sort out, but I wouldn't assume that the latter was either a copy or by WEM.

Osmund Bullock,

Mark, the point about WEM apparently not exhibiting after 1909 was not to imply he wasn't active after that date, but to suggest that there's probably no point in searching for the Evans portrait in the catalogues of the Grosvenor, New, Birmingham and Walker; nor, as Grant mentioned, in those of the RP and Modern Soc of PP, at which he seems not to have exhibited at all.

As to the likelihood of picking the wrong artist, I suspect that both men would have been pretty obscure by 1955, over a quarter of a century after they were (for different reasons) last apparently active. But I agree that we're unlikely to get a final answer unless the Dickinson archives (or the Frogmore painting itself) reveal it.