No sooner had the dust settled on Munich`s grand Islamic exhibition in October 1910,than another major show,this time devoted entirely to carpets,opened in November of that year in New York.It was curated by W.R Valentiner,head of the Decorative Arts Department at The Metropolitan Museum,who eventually became director at the Detroit Institute of Arts.Born in Germany,he was a protégé of Wilhelm von Bode,who actually recommended him for the position.The previous year he had showcased Dutch painters at the Met,which had been a huge success,so he was amply equipped to create the first major show of historical carpets in America.He seems to have been something of a rug-man,perhaps due to Bode`s influence.Curiously,the Munich exhibition is not mentioned in the catalogue`s introduction,although Valentiner surely knew of it.
The whistles of rug-afficionados in the US had already been wetted in January 1910 by the auction of the phenomenal Yerkes Collection,but Valentiner`s aim was to educate the public in an objective way,without the distraction of mercantile interests.Resources were garnered from ten major collectors,with fifteen pieces alone from the Lees Williams Collection.
The Kaiser Friedrich Museum was supposed to have lent the Berlin "Ming" carpet,but the promise evaporated despite the item `s publication in the catalogue .Why the loan was refused is unclear,considering Valentiner`s close connection to Bode,who purchased the rug in the roman art trade in 1886.However it had only lain four years in the Berlin museum as a purchase from Bode.The carpet itself is said to depict a Dragon and Phoenix duel,or was at least so interpreted by Bode,who was excellent at dreaming up monikers for old carpets("chintamani" is another cracker)but even Erdmann could not be sure of its appositeness.It certainly does depict two big birds,whose eyes form the focus of the rug itself(a "Nazar")The Phoenix,if it is such a one,has been obviously copied and re-copied into total obscurity.Erdmann suggested the carpet may have been much larger,and a simple mock-up demonstrates how this unlikely scenario may have looked. One interesting point is that at some time after its publication in Sarre-Trenkwald the carpet was restored and its major damages refurbished.Precise details of the restoration work are lacking,but nowadays most reproductions show the piece in this condition.Often mentioned in a breath with a painting dating from the 1440`s by di Bartolo,the Ming carpet was radio carbonised in 1999 with a likely date of between 1486 and 1645.It was also the victim of a fake attack in the 1980`s.
The Berlin piece shares a common technical feature with another early rug now in Stockholm,the "Marby" rug,so-called because it was said to have been found in a Swedish Church of that name,although it may have actually been acquired from a private source.On the back of both rugs knots have been inserted at approximately 5-6 centimetres distance.The threads were left long.A number of hypotheses have been advanced for this:that it was a knot count for a day`s work,or functioned as extra insulation or even a grip underlay.Such "back-threads" are also known from Abbasid work and Vase-carpets.The colour differs from the Berlin piece significantly.From recent photos the ground in the large octagons appears to be camel coloured,whereas the Berlin rug has a Konya yellow.The camel colour may indicate a production area east of Konya,unless it be oxidised natural white:most reproductions are too yellow. One closeup photo seems to convey the colour quite well(most reproductions do not even look like rugs)The Marby has been c-dated to between 1300 and 1420.The strips attached to its side were even older:900-1100.
|An Abbasid fragment with backshag|
The mundane Marby rug lacks the the visionary quality of the Berlin piece,with its writhing,snake-like contortions,but two further fragments should be considered.One item found at Fostat is now in the Museum der Kulturen,Basel,and depicts a Pars pro toto image of the Dragon`s back.C-dated to 1452-1635.17 x 32 cms.
The other item was first published by Erdmann in 1955.It was sold at Sothebys on 17 September 1992(lot 29) for $68,750. A Fostat foundling,It entered the Kirchheim Collection and was last seen in an Italian Collection,presumably the Bruschettini.The fragment shows a phoenix head and a part of the dragon in combat.A human figure,legs and arms akimbo,can be made out in the surrounding green field.As with the Berlin example,Erdmann suggested a double-rowed format,which with the extra wide border may not be far-fetched:it would also resemble the depiction in the di Bartolo painting. C-dated to 1395-1442.66 x 48 cms.
Proceeding thematically,Valentiner follows on with a salvo of Dragon carpets,all of which now reside in the Philadelphia Museum of Art.
The Reverend Williams`Dragon carpet is a strange thing indeed:Persian knotted to the left on a cotton foundation,and woven upside down,i.e the weaving began at the top end.The last 12 inches of the carpet(the beginning of the photo)was woven from the other direction and meets in the field.Perhaps the carpet had been removed from the loom and then re-mounted.That is,if C.G Ellis is to believed:or the top-end(here the beginning) is a re-weave.The design is a simplified version lacking the deer and fawns,but still employing the alternating Dragons and Lion/Kylin depiction.Valentiner`s comparison with the Berlin and Lamm rugs is inaccurate,as there really is no similar ccarpet.
A spacious and attractive carpet donated to the Philadelpia Museum of Art by the collector Phillip Sharples.With repeating Dragon design,but lacking Lions and Kylins.The center forms a large,half-concealed medallion.C.G Ellis related it to a number of other rugs,three of which are shown here.
|Philadelphia Museum of Art|
John McIlhenny was inspired by Lees Williams to collect carpets.The last of the three Dragon rugs is a member of a later simplified group of which the following is a streamlined example.Charles Ellis called the large Sickle-Leaves which form the compartments"the General`s Moustache"At least eight examples from this group have been published,including two examples in Turkey on a blue ground.They form an intermediary stage between the repeating Dragon design and the final"dragonless-Dragon" group.
|Philadelphia Museum of Art|
|A "dragonless-Dragon" carpet-Hinrichsen-Bausback|
A seemingly old Caucasian carpet in the "Portuguese" style,aping a small group of Persian or Indian carpets from the 17th century.It forms a group with a carpet in the TM(the DeMotte) and a truly old-looking fragment in the Orient Stars Collection,ex-Spuhler.These three items all feature a square,stepped central medallion,unlike the originals and two other rugs,the Balfour carpet and a damaged and cut rug in Berlin,which are design-wise closer to the originals.17th century items sometimes feature black and Indio slaves.
A red-ground 18th century Caucasian carpet with addorsed Sickle-Leaves and dotted dragon remnants.An interesting rug which seems never to have been published in colour;even the Met`s black and white photo is poor.For comparison,a simpler example once with Eskenazi.
A green-ground Small Pattern Holbein rug from the Joseph Lees Williams Collection,now Philadelphia (55-65-5)C.G Ellis was of the opinion that this carpet hailed from Wallachia,Romania,or possibly central Anatolia,and he dated it to the 17th century.In a portrait of the young Edward Vi he claimed to have found a very similar rug.But in virtually all details it is different.Not even the border appears to be the same.Ellis pulls in a carpet from Sivrihisar now in Istanbul as an example with such a border,but that carpet makes a poor attempt at reciprocity.The silhouette is much clearer in 19th century revivalist rugs from West Anatolia,particularly from the Bergama / Yuntdag area,where it was employed frequently.Further discussion:HERE Nrs 204-208.
Monumental in scope,yet very elegant...see HERE 046 et al.
Now in Philadelphia.
Another green-ground Holbein rug,but of genuine age,showing up plate 7`s iniquities.Further discussion: HERE
Now in Philadelphia.
A so-called "Chessboard Carpet",further discussion: HERE
Now in Philadelphia.
A small bird rug of the type which fathered countless fakes.Attributed by C.G Ellis unjustifiably to European Turkey.Further discussion HERE
Full discussion: HERE
Now in Philadelphia.
Anatolian type Lotto rugs with cloudband borders are uncommon.Due to its excellent condtion Ellis thought the following Lees Williams rug could have been 19th century,and yet he inconsequently dated Plate 7 to the 17th.Two similar examples are shown here for comparison.Further discussion: HERE
Another unusual carpet from the Davis Collection,which was rich in such things.Rows of Turkmen-like shrubs alternate with Latch-hook göls.The shrubs have been inverted at the top.Very pale,pastel colours,reminiscent of 19th century Avar work.Shown as usual upside-down.Seems never to have been published elsewhere.
The Davis proto-Lenkoran carpet,a type which never went into series,but paves the way for the later 19th century items based on the Dragon carpet heraldic shield with rampant Wyvens.Dragon carpet border.
The only known complete example of this rare Ushak group.Full discussion:Here
A spacious Star Ushak carpet,now in Philadelphia.A similar layout can be observed in the Benguiat/McMullan example,now in the MET.
An Ushak medallion carpet combining Lobed and Star octagrams,as can also be seen on a fragment in the Kirchheim Collection.
An otherwise unpublished example of the Caucasian Shield Group.Not mentioned in Pinner/Franses,Hali I-1.The human figure is unknown in such carpets.A group of Turkmen asmalyks,attributed to the Tekke,employ a similar border,even down to the "kink".
An unusual use of the roundels normally found in the borders of Mamluk carpets to create a central medallion.Note the Lotus forms in the border-cartouches.
Sold for $785,000 at the Clark Sale,Sothebys New York,5 June 2013,Lot 1.
Further discussion: HERE
This sad survivor of a mighty group has been attributed to Hamadan or Kurdistan on the basis of a statement made by A.Cecil Edwards,an unreliable source.Equally dubious are the attributions to Khorasan on the basis of structure,i.e the Jufti knot.In fact no one knows from whence they came.This battered fragment was also shown at the MMK Exhibition in Munich,and had to be shipped across the Atlantic to New York in less than two weeks.A reconstruction shows the approximate deployment of the various pieces.Such carpets were made with two basic borders:a Strapwork type often associated with the "Golden Triangle" group,and a meander-leaf,such as here.An exception in every way is the Keir Collection example.Even modern copies are known,from the Hamadan/Malayer district.As a whole,the group has something quite regal about it-The Wealth of Kings.
|Some early 20th century carpets|
Cut,worn and disfigured...the Williams Tree carpet was another transatlantic tourist at the Munich Exhibition two weeks prior to the opening in New York.Plate 49 in the MMK catalogue,further discussion: HERE
Now in Philadelphia.
A large medallion formed out of concentric circles of cartouches and plaques on a whiteground arabesque border.The vertical thrust with a Dragon and Phoenix depiction became,in a simplified form,a border device.Or was it the other way round?Dr.Thompson considered this carpet to have originated during the Timurid dynasty.Apparently this example is in better condition than the pair in Lyons.Ex-Yerkes collection,now MET.
Said to have been brought from the Ardabil Shrine,itinerant rug-fans could have seen both matching pieces in Munich and New York within a short space of time in 1910.Although the Sarre/Doha carpet was not published in the main catalogue,it did appear in Volume IV.It appears to be in much better shape than the New York piece,now in the MET.It seems old Mr. Yerkes was often sold mutton dressed up as lamb;but then again,he rarely inspected what he bought.The Sarre/Doha carpet seems to have survived well, if a comparison between a 1908 and a contemporary reproduction is anything to go by.
The Baker Salting Carpet,now in the MET.The "al-Khidr" carpet is one of four large Salting rugs.
An Animal-Medallion carpet from the Williams Collection,now in Philadelphia,of a type which rarely utilises such an extraordinary arabesque medallion.Only the Ardabil and Anhalt carpets are superior in this respect(and,one might add,the field of a large fragment in the Musee des Gobelins).Yet a coarse version of the banded medallions found on the Sanguszko carpets is used for the corner-pieces.No other carpet of this type is known.
The controversial Pratt fragments,ex Bardini-Yerkes,now in the Brooklyn Museum.(2-the left illustration is a photo;the right a painting from the Yerkes catalogue)Purchased at the Yerkes Sale in 1910 by Cottier & Co for $5,600.
Erdmann thought they were part of the missing Hatvany fragment;Ellis and Thompson did not.
In the 1931 photograph published by Friedrich Sarre(3) the main border has been placed on the left side,but by 1936 it had been sewn onto the right,as can be seen in a 1936 Hungarian rug catalogue(3A).This can also just be made out in a photo of the fragment in Hatvanys` study,framed and behind glass(4)The carpet itself is said to have come from a Polish castle.
The Pratt border fragments(5) and the Hatvani separated border,plus a fragment in Lyons(6) and another from the Schorsher/Wher collection(7),are,in this author`s eyes,from the same carpet.The cloudband-scroll outer border can be seen on the upper Pratt fragment,which bears the inscription(8).The upper border also shows a semi-cartouche with Kylins.A part of such has been used to plug a hole on the left-hand side of the Brooklyn fragments(9).The border of the Hatvany fragment is identical.Both the main Hatvany fragment and its associated border have been patched with what appears to be fragments from a Kashmir shawl(confirmed by Sarre)-which means they were repaired at the same time,although it is still not an ultimate proof that they originally belonged together.Such creative restoration work was typical of the Bardini workshop,and we know the Yerkes-Pratt-Brooklyn pieces were from that source.Perhaps Bardini was also the seller of the Hatvany fragment,which is not unlikely(Polish castle or not)
This does not automatically mean the Pratt fragments are part of the Hatvany,as we cannot be sure that the fragment sewn onto the main fragment(with medallion) is actually part of the original,although Sarre`did not question this.It is not clear from Sarre`s report if he had actually handled the Hatvany fragment.However,the carpet closest to the Hatvany is the Paris-Cracow,and a study of the two pieces is helpful(10).Both have a similar treatment of flowering trees on a white ground(which,however,the lower fragment of the Brooklyn fragments does not have)Both have medallions etched with flowers;both feature the small yellow-ground minor guard with crosses;both have an inner black-ground border;and whilst the outer border guard is different,it is a meander.Both also feature Kylins in the main border.The overall drawing is similar.
If we take it as given that the Brooklyn border fragments(and the Lyons and Schorscher/Wher collection items) are from the same carpet,and that the border sewn onto the Hatvany is original(which Sarre did not doubt),then we can say that all these medallion-cartouche borders are from the Hatvany.Otherwise they are from a completely unknown carpet which has perished.But taking what we have as a given,then they are all from the Hatvany fragment.
This leaves us with the main segment of the Brooklyn fragments,the "Angel" corner section,and the lower yellow/white ground fragment of a field.Ellis felt that the drafting of the Angel fragment was too different to the Hatvany medallion to have "started life together",although that could be said of the Bode/Getty Crane pair which feature a similar composition with corner-piece Houris(11).
The Hatvany has the unusual feature of a pendant abutting into the field on the left side of the medallion(12,13).This is not a quarter corner medallion as Jon Thompson thought,but just what it looks like:an extra pendant,which must have mirrored on the right side too(14).This device is shared by a number of Sanguszko carpets(the Miho,Taylor,V&A and Tehran examples)The actual corner piece would have started further up the field,beyond the medallion`s pendant,and is not visible in the fragment.The yellow ground cross minor guard is separated on the outside main borders by a single black line.On the guard adjoining the main field on the Hatvany fragment it is a red line-just as on the Angel fragment,at the top.Thus the Angel fragment could have lain within the field`s radius.
As for the yellow/white ground fragment at the base of the Brooklyn fragments(15),it seems to be from another carpet than the Hatvany.As Jon Thompson said,the Hatvany has no sworls,and the Brooklyn fragment no flowering plants.The drawing of the Brooklyn fragment is generally acknowledged to be superior.
So,we have the following equation for the enigma of the Hatvany and Yerkes-Pratt-Brooklyn fragment:
Brooklyn piece-side borders and inscription panel:probably part of the Hatvany fragment,IF the border sewn onto the Budapest piece was original.
Brooklyn piece-Angel fragment:could be part of the Hatvany,60-40 chance.
Brooklyn piece,-yellow/white ground, lower field panel and triangular patch:probably not part of the Hatvany fragment.
There are no published photos of the back of the Brooklyn fragments.It would be interesting to see if the weave is identical in all fragments;or if the side and top border are identical,but different to the Angel segment and the yellow/white field panel.And if the Angel and yellow/white field panel differ.This could be ascertained in Situ by an experienced examiner and would help determine that which can be determined,failing the re-appearance of the legendary Hatvany fragment.
A further discussion of the negelected inscription on the Brooklyn fragments can be seen HERE
|3A-The Hatvany with border sewn on the wrong side|
One of three Kashan silk rugs from the Altman Collection,all of which went to the MET.There are ten known medallion carpets(excluding 3 with animals)The consonant to this is the Goupil carpet in Paris,perhaps the most beautiful of all.The Altman piece was later published in Sarre-Trenkwald( II-41)
With so much superior material on offer at the Yerkes Sale,this was an odd choice for the Metropolitan to purchase.Certainly an oddball carpet with its Garden Carpet corners and more than a hint of India.If not Indian,it certainly inspired a series of later copies:
A Salting-style carpet from the Altman Collection,of which no decent colour photo exists.A closeup must suffice,and a good black and white reproduction from Sarre-Trenkwald.The Metropolitan seems not to have bothered much with this item;perhaps it is a dubious quality.Sarre-Trenkwald give cotton warp and silk weft,but there is no mention of metal-thread,which however appears in the closeup surrounding the split arabesques.The piece is of a group with Vase and Trees,as in the Topkapi,the Dubroff carpet and a rug illustrated by Reed.Why the Altman rug has been neglected is unclear,but it deserves to be investigated.
The Cochran Animal-medallion rug,one of a pair,both now in the Metropolitan Museum.A simple identification of the two rugs can be seen in the different treatment of the rams on both sides of the border.The Yerkes rug depicts the ram`s legs thrust forward,straight out;in the Cochran the legs are elegantly kicking in the air.In fact the two pieces were confused in both Martin and Sarre-Trenkwald.The rug illustrated by Valentiner is the Bardini-Fletcher-Cochran rug;that in Martin and Sarre-Trenkwald is the Capponi-Yerkes-Fletcher.In this case Mr.Yerkes actually got the better deal:horrible things have been done to the Cochran,and it is usually not displayed.But the Yerkes-Fletcher is a virtual copy with a medallion of jovial musicians shown in quartered mirror-image.
|The Cochran ram|
|The Yerkes-Fletcher ram|
Subsequently published by Hawley and Dilley,this was exhibited at the MET`s loan exhibition of "Polish" rugs in 1930(Acc.10.103).It has subsequently disappeared completely(as have other rugs)from the Museum`s collection.A search on the MET`s website produced no result.A red-ground rug with a blue-green border,it is not unlike another rug in the MET from the Fletcher collection(Acc Nr.17.120.143)
A tired looking rug from the Clark Collection which managed to fetch 74,500 GBP at Christies trial run of pieces from the Corcoran Gallery on 24 November 2009,lot 207.Spuhler 102.
From the collection of the financier General Brayton Ives,who purchased it at the Yerkes sale for $4700.Thence to Senator Clark.It was later sold at the Sothebys Clark/Corcoran Sale on 5 June 2013,lot 7,for $173,000.A pair to this was offered at the Palais Galliera Sale(Rothschild Collection) on 28 March 1968.(Spuhler 165)
|Sothebys Corcoran/Clark sale 2013|
The second Polonaise rug from the Ives-Clark connection,sold at the great Clark Sale by Sothebys on 5 June 2013,lot 9,for $ 137,000.
Poorly documented on the MET`s website.Later published by Pope,and by Dimand.
The most expensive rug in the Denman Waldo Ross Collection,later given to the Boston Museum:Ross paid $7,500 for it in 1907,purchasing from Dikran Kelekian in Paris.
A rare large Polonaise carpet from the Widener Collection,now in the National Gallery,Washington.For compaison:the Henry Skirmunt carpet.
|Count Henry Skirmunt|
The second Widener Polonaise,now also in the NGA,Washington.According to F.Spuhler(105),purchased from Rothschild.
One of a group of carpets decorated with large arabesque bands,the Clark/Corcoran did not always look so poorly-see Erdmann`s illustration from 1961.A number of half-way intact pieces,and three other morceaux have lately come to light.Sold at Christies 24 November 2009,lot 30,for $100,900.
The following four items are from the "Red-ground Floral" group,a type of rug which developed out of the Emperor`s carpet,as discussed HERE
The Clark RGF("Red-Ground Floral") carpet sold at Sothebys on 5 June 2013 for $365,000(lot 23)This was a Bardini carpet,which was later sold at Christies London in 1899.RGF carpets with a cartouche border are rare and usually in smaller sizes,although at least two large examples are known(The Tucher and another piece in the Corcoran)A similar example is in the Al-Sabah Collection,Kuwait.The rug is shown here together with the following item in the catalogue,an Indian version on cotton ground(the Persian original has a silk weft,although Sothebys analysis of a woolen weft is hard to believe)The Mughal rug(lot 24 in the Clark sale)brought marginally less-$365,000.The difference between the two can be seen in the robust austerity of the Persian compared to the sweetness of the Indian version.
An RGF carpet from the Clark collection,one of a small group with metal thread.A simplified version,somewhat awkward by comparison,of the Enzenberg-Thyssen example,which features elegantly drawn swooping birds in field and border.Sold for $ 245,000 at the Clark Sale(lot 6)
The Widener RGF later appeared twice at auction in our time.At Sothebys on 12 April 2000 it brought $159,080;and at Christies seven years late on 25 October $148,555.A smaller version with similar design sold at Sothebys in 1997 for $310,500.
|Sothebys 10 April 1997,lot 114|
A lofty and elegant RGF with sickle-leaf design,now in the MET.There are a number of Indian versions of this design,recognisable through their use of racemes.Three other pieces are shown here for comparison.l-r:the Altman-Lourical Portugal-Thyssen-Sothebys-Tehran carpet Museum.
The Davis garden carpet is half of a piece divided between the Tehran carpet Museum and the Jim Dixon Collection.The Dixon fragment appeared in 1927 at the AAA sale of Alphonse Khan;the Tehran piece was at Sothebys on 14 April 1976(lot 5)
Fragment from an Imperial grade Indian carpet,Pashmina wool on a silk foundation.Scan 48 shows its current conserved condition(the smallest fragment is from the Friedsam Collection)Scan 49 gives an idea as to which scale these weavers worked;Scan 50 shows the Altman piece,then and now.Scan 51 brings the other known fragments together.Such premium rugs feature a spiral-vine within the lattice.
A silk pile carpet on cotton foundation said to originate from the Church of Apportas in Portugal.Sold at the Clark sale in 2013 for $545,000.It has a twin which first appeared at Sothebys in 1987 and was sold for $32,205.At a return sale with Christies in 2019 it reached 243,750 GBP.Four long rugs are known,including the Cassirer/Detroit piece now in the David Collection.A piece divided into two and sold to Kuwait and Qatar was still together in 1936, as a photo from the MAK Library ,Vienna,shows.Four other rugs do not quite fit:the Dixon fragment and a carpet at Sothebys in 2018 both have a different border and may well be later copies;the carpet from two Benguiat sales is not clearly confirmed as made of silk;and a piece from the 1933 AAA auction cannot be placed due to lack of other examples and its poor reproduction.
With the Boston Museum since 1893 when it was donated by the Ames family,this Indian pictorial rug had been purchased in London as early as 1882.The central theme is a gaja-simha,a mythic combination of lion and elephant,being attacked by a simurgh.As in all Indian animal carpets,creatures flit about and chase each other;there is very little actual combat.
Once with the Duke of Rutland and now in the National Gallery Washington,the Widener Animal Carpet with its one-directional narrative has been deeply influenced by the miniature painters of its time,and features a menagerie of wild animals in chaotic flight centered around a mahout driving his elephant.A leopard standing upon a dhurrie-covered cart is being chastised by a swordsman with a fly-whisk.From a border of blue quatrefoils and white cartouches a pink partridge flys or stands guard; a grotesque face stares out at us.
On January 15 the exhibition was dismantled and the carpets returned to their owners.Whilst W.R Valentiner was unable to scale the lofty heights attained by the Munich exhibition,the show had been strong in the horizontal,packing 50 great rugs into just three rooms compared to Munich`s eighty.The Great War that followed led to a partial dissolution of the Hapsburg holdings,and the acquistion therefrom of one half of the Emperor´s Carpet,arguably the most influential carpet ever woven.The MET had to wait a long time for another Carpet show equal to the 1910 event,but in the meantime was blessed with so many gifts and purchases that it became one of the world`s great repositories of Oriental Carpets.