Friday 3 January 2014

The Holbein-Lotto Family II:The Large Pattern Holbeins.

A Group of Coptic Textiles

Kurt Zipper was the first to point out the similarities between a group of Coptic textiles and the Holbein-Salor Göl,in an article in the magazine "Heimtex" from 1980.The idea was further developed by Peter Trimbacher in his "Updated History".There are more explanations for this:Turks settled in Egypt in the 4-7th century and influenced the local weavers;products from Central Asia were exported to Egypt;or the Turks themselves copied these designs from Egyptian originals which they found on entering Anatolia.The third possibility is more plausible.These are the oldest known textiles with this design.


41-Zipper,Heimtex 9,1980(2)



44-Rietberg Museum(53)

Acquired in Cairo,this fragment from the Keir Collection features a minor Göl from an LPH carpet

45-Fustat fragment,Keir Collection

The justly famous Benaki fragment in Athens has been mooted as a Timurid production

46-Timurid(?)fragment,Benaki Museum

47-A reconstruction of the above

A carpet from Beyshehir(Nr.12)with an archaic look but highly simplified "Göls"

48-Gantzhorn 249

49-Beyshehir 12-still in Konya

And the famous Bernheimer chair

50-Milestones 30

The Para-Mamluke Carpets.

From the Lees Williams Collection in Philadelphia,an early,simpler version of the 4-and-1 design found on Mamluke carpets.Contains both Persian and Turkish knotting,and has been attributed to Anatolia,Syria and Egypt,although the current tendency is to place it in East Anatolia or even Western Persia

51-Ellis,Philadelphia 1

The prayer carpet from the Chehel Sutun Kiosk in Isfahan,a self-conscious work in the manner of a "Turkish Salting" carpet.Its laboured design and inscription have led some authorities to doubt its authenticity.It is finely Turkish-knotted. (analysis P.R.J Ford,Hali 42-42).However,a similar rug appears in a painting by Giovanni Udine from 1501,as discovered by John Mills(see Hali 93,page 73)The only carpet depicted here with an elaborate inscription(Hasten to repent before death)Now in the Carpet Museum Tehran,but how did it get to Isfahan,and when?

52-Gantzhorn 215

The latest discovery in this field is a carpet once in the London trade and now in Doha(see Thompson,Milestones page 145)


Two Soumak woven carpets have been placed here for convenience in the "Para-Mamluk" Group. The first example was the starting point for C.G Ellis` famous article concerning the "International" style.The second carpet,now restored,is in the Hermitage Museum in Leningrad.There is some disagreement concerning the two rugs:Jon Thompson believed he could see original knotting in the TM piece, the hypothesis being that it is a very old knotted rug which has been restored with Soumak technique.Tsareva(Hali 154-96)dismisses this.The two carpets are remarkably similar.It`s hard to believe that this is 19th century Caucasian production as the scale is truly epic.

54-Textile Museum,Hali 154-98

55-St.Petersburg Hali 154-96

A delightful roundel of Para-Mamluk in the Kunstgewerbe Museum Dresden was the subject of an essay by C.G Ellis in Hali 93.He describes his surprise at realising that Lessing`s reproduction from 1877 was a fanciful reconstruction based on a painting by Anguissola,from 1555.This fragment is Turkish knotted.

56-Berlin Hali 93-76-118

57-Lessing 1877

Finally,a fragment once in the Bernheimer Collection,asymmetrically knotted,and with an early form of the curled-leaf border

58-Berlin Hali 71-119

The Large Pattern Holbein Group.

1 Medallion Group.

With budded Kufi "F" border,this was the subject of a spirited article by Robert Pinner in Hali 63-69,in which he compared the inner latch-hook medallion with that of the "Eagle-Göl" Group

59-Iten-Maritz,Turkish Carpets (43)

An early sign of corruption on a carpet found at Divrigi-without the node appendages this would no longer be recognizable as an LPH carpet

60-Divrigi Vak 15

2 Medallion Group.

Nicknamed "Domes and Squinches" in a legendary Hali review by John Eskenazi(H.32-9)is a carpet in Istanbul whose medallion centres recall the Lees William Para-Mamluk.The cornerpieces resemble large butterflies or moths,which is almost certainly the way the painter Beccaruzzi saw it in the 16th Century

61-Beccaruzzi,Portrait,Uffizi Gallery,Florence

62-Domes and Squinches,Vakiflar Museum Istanbul


A companion carpet in the Museum of Islamic art Cairo has been published for the first time in colour in Milestones.It first appeared in 1949,many years before the previous rug was discovered at Divrigi.Dismissed by C.G Ellis as "late",it has now been fully rehabilitated by Jon Thompson.It shares the field division of the LPH carpets in Berlin and Mailand.This results in a heavier layout than the Divrigi rug,but the rotation of the inner medallions is even more pronounced,with the whirligig stars repeated in the field.

64-Milestones.figure 142,page 148

Here for comparison are the three stars and bars medallions




A carpet from the Church of Sion,discovered by the late Marino Dall `Oglio, has a whirligig centre and an F type budded Kufi border


Another group of Two-Medallion rugs, closely related.

The carpet in Berlin,once with Bode,has an especially bright white ground accentuated by its missing outer borders.The concept is that of the Divrigi"Domes and Squinches" carpet,an endless flow of repeating major and minor Göls.The corner-pieces when placed together form a new minor Göl,not unlike,but less elaborate than,the secondary on the Salvadori-Keir carpet.


A similar variation occurs as the main Göl on a carpet  once with Chris Alexander

70-Alexander Collection 197

The Berlin carpet has been simplified and the treatment spruced up.A "black-and-blue"outer border makes an appearance here and is reprised in 19th century Caucasian carpets.Overall,there is more figure-ground ambiguity,but no further use of the "Stars-and-Bars".

71-Berlin-Beselin 73

The other Two-Medallion carpets here are fenced off behind borders in the manner of a picture frame.Even the Cairo carpet suggests this.

Two carpets are from the Sey Baba Yusuf Mosque in Sivrihisar.The first has an unambiguous medallion design.This model re-appears in Bergama carpets of the 19th century.The same border is seen in a painting by the Master of Saint Giles from c.1500

72-Istanbul ICOC 2-14

73-Master of Saint Gilles(the painter is perhaps the cleric holding back the green curtain)

In the second rug from Sivrihisar, the figure-ground ambiguity has been preserved,creating a device often seen on Ushak carpets(upper panel)


A third example is from the Keykubat Shrine in Konya,where the last Seljuk fragments were found.With some vestiges of a Kufesque border


A piece from Kent-Costikyan was exhibited at the 1926 Chicago Exhibition,where it was praised as superior to the Berlin example(which it   does not resemble)Probably missing its outer borders.


A further six pieces are worthy of mention,and can be downloaded from Flickr:a very pretty fragment in Berlin,purchased in 1904,dated by Dr.Spuhler to the 19th century;a Konya carpet sold at Lefevre`s in October 1979 for 1500 Pounds,described as Bergama;a piece published by Denny in his Classical Anatolian Carpets book and later sold at Christies on 5 April 2011(50)for 43,250 Pounds;a carpet published by Ledacs with a debased design,although still showing Stars-and-Bars;an old  piece showing further signs of design weakness,from the Mosque at Sivrihisar,shown at the Istanbul ICOC;and a carpet from the Orient Stars Collection which mixes the "Crevelli" medallion design with the LPH Holbein.

Download  HERE

3 Medallions

The four pieces in this category represent a further simplification when compared to the foregoing,and certainly when compared with the large Multi-Gol pieces in Berlin and Mailand.

A carpet in Philadelphia,once with Friedrich Sarre,still presents the"Stars and Bars" design of the earliest International style,but the borders employ a simple Kufesque,and the corner brackets are highly schematised.Notwithstanding,this is certainly one of the prominent examples of an LPH carpet.

77-Philadelphia 2

The "Wheel" Carpets,as they were called in old inventories,were also copied in Spain

78-Alexander Collection 147

Berlin 1 has splendid whirligigs in the medallion`s centre and a simple C1 border.All three of the next rugs employ a meander type"Bardini" border,also seen on some Lottto carpets.

79-Berlin H-154-123

Acquired by Mr.Myers in 1928 and now in the Textile Museum,in the following rug the weavers appear to have run out of green in the border,but it may be that this sea-change was necessitated by the lack of visual impact:the green meander line disappears on the green ground of the border.The utilised salmon colour is hardly better.The three medallions are identical.

80-TM Denny 11

The Lotus border occurs again on the Two-Medallion ICOC carpet(St Gilles painting).Here the central medallion differs,a device used extensively in 19th century Caucasian rugs.

81-Spuhler 5

Multiple Medallions

The first of the following two carpets reigned alone for more than a 100 years until its counterpart appeared at an auction in Venice.The Berlin rug is less crowded,the Tabibnia carpet seems more imposing and is larger.Both share similar noble Kufi borders and Kochanak.The second carpet is fully compartmentalised,whereas in the Berlin carpet the elegant rosettes(which also appear on the Dirksen SPH)are distributed more freely.The secondary Göls on the second carpet are derived from the centres of the main medallions and fenced in with a Coptic surround later seen on Salor chovals.The Semenzato carpet is in very good condition having been careful preserved for 150 years by its previous owner.




The above foto is taken from the original catalogue,where it had been estimated at 5-7000 euro.Sold at Semenzato,29 November 2002(536) for $567,120.

Two further relevant pieces are in Turkey:a carpet from the Sey Baba Yusuf mosque in Sivrihisar with Stars-and-Bars medallions,mock-Kufi border and Salor minor Göls

85-ICOC Catalogue 2(21)

A carpet from the Vakflar Museum was shown at the Istanbul ICOC,and had also came from the Sivrihisar Mosque.With the same mock-Kufi border as the Two-Medallion carpet from Konya,and the above


Finally,a later decadent carpet with multiple medallions and vernacular border,taken from the Keykubat Mosque in Konya


Octagonal Medallion Carpets

The fragmented piece in the Bavarian National Museum Munich has been the subject of a number of reconstructions.It was originally purchased from Bernheimer.

88-Munich-Hali 169-25

89-ORR 8-4-30

90-Ellis,TM Journal 1-2-7

The stucco decoration from the mausoleum of Nur ben Ali in Uzgen,Kyrgyzstan dates from the early 11th century and probably later inspired the borders on the second example of this group,the so-called "Wind-Carpet"

91-Thompson,Iranian World page 50

However it is really quite close to the Berlin Museum`s allover yellow ground carpet


93-The Wind Carpet,Christies October 1994(519)

Apparently found on a Swiss flea market,the carpet was eventually sold for charity at Christies London on 20 October 1994(519)for $170,660(see Hali 78(127)It lacks the subsidiary Göls which would otherwise have taken it into the next group.

The 4 and 1 Group(or 2-1-2)

As with so many things the origins can be found in Egypt

94-Christies April1992(322)

The above sold in London for $57,950,later with Mikaelov,now in the  David Collection.

It seems a logical idea to simply curtail the further progression of the large Holbein carpets,thus resulting in a quincuncial design.The following iconic rug,exhibited many times world-over,has a border resembling that of Holbein`s "Ambassadors"


96-The Ambassadors

Berlin`s SPH is repeated here for its similar border

97-Hali 69-95

The Detroit carpet is a fascinating and neglected example with two unusual half-borders spliced together,and well-drawn Salor Göls

98-Detroit Institute of Arts,Gantzhorn 364

A well-travelled carpet at various times with John Eskenazi,the Textile Gallery and Franz Sailer, later sold at Christies NY on 12 September 1989(108) for $44,000.On a rare camelhair ground with Salor Göls and a nicely positioned central medallion which,however seems lackadaisical.Hanna Erdmann located it in Eastern Turkey,as did the author of  the APG in Hali 48(88)

99-Franses-Christies 12 September 1989-108

Alberto Boralevi modestly dated his piece to the 18th century,although it could be much older.With interesting corner boxes in the central medallion, Salor Göls and swastika border

100-Boralevi 14

Published only once with a substandard photo,this intruiging piece from Alanya measures 150 x 196 cms.The panelled minor Göls and cogwheel centre give it a Mamluke touch,the pendants above and below the central medallion recall the Salor

101-Alanya Gantzhorn  ill. 495

A fragment acquired for the Orient Stars Collection has a neatly drawn Salor Göl,but may be from a re-entry carpet

102-Hali 81-53

A carpet from Moshe Tabibnia reprises Berlin`s yellow-ground piece with plaited octagon rosettes

103-Milestones 4

104-Hali 148(29)

Despite differences in treatment the similarities to the foregoing piece are clear

105-THC 5-586

A simpler,more village-like version drifting slowly further away from the original design,but sustained by outstanding colour

106-TIEM(Gantzhorn 365)

A carpet with Stars-and-Bars once sold at the von Bode Estate Sale in 1929,now missing

107-Bode 69

In his description of the following item (Orient Stars Catalogue Nr.161)Friedrich Spuhler attempted to assign it to the Para Mamluk orbit,but without much success.The carpet is excessively symmetric,lacking any surprises.Perhaps a much later work,or over restored,or both.


First published by Grote-Hasenbalg,the next rug appeared at Nagels on 7 December 1977,later McMullan Collection.


A carpet from the Parsons Todd Collection with flaccid layout but archaic border

110-Todd Hali 64(128)

In the following carpet from Franz Sailer the nodes of the central medallion have been replaced by more easily-drawn latch hooks.The borders are a charming village rendition of the C Type Kufi border

111-Sailer Sale 33

Compelling in its awfulness,this rug went unsold at Christies on the 12 December 2006(64)against an estimate of $3-5000.It had previously appeared at Sothebys on May 30 1987(19),estimated at $1000-1500.


A final carpet in this section was purchased by Heinrich Kirchheim at Rippon-Boswell`s on 19 November 2005(10) for $6,185,after having failed to sell at Nagels on 14 November 1997 against an estimate of DM 6,200.Probably East Anatolian,19th century.

113-RB 2005

The LPH carpets continued to exert an influence on carpet-making well into the 19th century,with a number of interesting examples from the 17th and 18th.Those lay outside the scope of this article but may well find their way into a future entry.

Part III

The Holbein-Lotto Family

The term "Holbein" makes its first print appearance in Werner Grote-Hasenbalg`s book of 1922.In 1935 Kurt Erdmann was still using it as a catch-all phrase to include the Lotto carpets. The names are inaccurate but have been around so long that it is probably better to keep them.Before examining the different groups,an effort will be made to list the existing types of Kufi borders.This was first attempted by Robert Pinner and Jackie Stanger in Hali 1-4(326)Both Holbein and Lotto carpets have been counted together,thus enabling  a view of the overlap between the two types.The “Seljuk” carpets with such borders have been dealt with elsewhere.(  Seljuk ) A monumental carpet discovered at Divrigi is an exception,two “Bellini”prayer rugs also have a Kufi border,and the Para-Mamluk Group will be dealt with separately.The Small Pattern Holbein carpets(“SPH")only use the Kufi border,apart from one carpet in Berlin.The Lotto carpets employ six Kufi variations.Others,as we shall see,were never used.The Lottos became increasingly adventurous in their choice of border design.

Type A Border.21 examples,19 SPH and 2 Lottos.Code-Name "Chainlink"



Type B Border.(Pinner Type C)23 examples,21(one fragment exists without field) SPH and 2 Lottos.Codename "sans chainlink"




The above shows the evolution of type A and B.The B type has lost the interlaced form and been replaced by a simple cross.

Type C Border.23 examples, 22 used only on Lotto carpets,and on the Bellini Prayer rug in Berlin.Code Name "Haring"



8-After Lessing

Scissor forms appear between the box elements,replacing the cross form.The double track support lines of the A and B borders have been replaced by a single interlacing line.

The line always runs around the inside of the border,except in the case of a fragment in the MacMullan collection,where the line passes along the outside


Type C1 Border.5 examples,3 Holbeins and 2 Lottos.This is the variation of the C Border used on Holbein carpets.Instead of the "scissor" shape,a double-ended volute appears,and the individual units are no longer connected by a single line.

10-From the Mediash Holbein

11-From the Philadelphia Lotto,Ellis Nr.7

Type D Border.(Pinner Type C2)Codename X-Box.12 examples,6 SPH,6 Lotto,+ 1 border fragment.


Type D1 Border, 3 examples,all SPH.This is a still connected form of the D Border.


Type E Border.Codename "4 Arrows"14 examples,12 Lottos,1 fragment and 1 SPH,plus one Bellini prayer rug and one carpet with "Lesghi" design.



A carpet in the Detroit Institute of Arts has not beeen fully investigated.Its true age is hard to estimate,but it appears to be the mother of all Lesghi-Star rugs.The star form appears to be a mutated Holbein Octagon

16-Gantzhorn ill.350

Type F Border.Pinner Type D+D2.Code name "Kochanak"These feature an extra border as seen on Turkmen carpets.There are three variations

17-The Budded Kufi

18-C Type -Haring Kochanak

19-Kufi Flower Kochanak

Type G Border.Pinner type A&B.2 examples,both Holbeins.Also used on 3 Para-Mamluk carpets.This is the noblest Kufi form,as depicted by Mantegna


  Jon Thompson has traced this design back to 14th century Persian Jalayarid paintings.(see Milestones)

Type H Border.4 examples.This rare border was the subject of an interesting reconstruction by Christina Klose(Carpets in the Iranian World,page 77)Sometimes called the "Timurid" border



Large Pattern Holbein(LPH) Kufesque Borders.8 examples.


The Divrigi carpet was discovered by Walter Denny and G.G Ellis in the Mosque of the town.Has been C-dated to the 14th century.

The two LPH("Large Pattern Holbein") carpets from Divrik and Cairo have similar Kufi borders,symmetrical in all directions



A two-medallion rug from Sion and another single medallion carpet in Istanbul have the F Border (budded Kufi)



The three wheeled LPH in Philadelphia uses a coarsened F Border Kufi Flower


and an LPH carpet in Istanbul shows a decorative move away from the epigraphic,albeit with interlacings


The Para-Mamluk carpets.

These reveal 3 types of border spread over  7 carpets.

The Chehel Sutun Prayer Carpet,the Doha Carpet and the Milestones fragment all employ the G Border

30-Chehel Sutun-Ganzhorn 215

The Philadelphia and Ellis/Wher/Tabibnia examples have an elaborate C Border with more sophisticated interlacings


Two Soumaks in the Textile Museum and the Hermitage will be placed,for convenience,in the Para-Mamluk category


The above is a variation on the G Border(Mantegna)but with interlaced double tracks.It appears only on the two soumaks and in an Altar  painting from 1503 by Francesco de Morone featuring an LPH



One solitary Lotto carpet is now in the Jaquemart-Andre Museum Paris and represents a simplification of the Kufi Flower Kochanak F Border


A similar border was illustrated by Lessing(Nr.20)where it is said to be located in the Church of Nazaro e Celso,dating to the late 16th Century


The well-known Mamluke fragment in the V&A Museum,London also features a variant of the G border.Original or pirate copy?

36-V&A,Gantzhorn 204

Two "Bellini" style prayer rugs also feature Kufesque borders. The famous Berlin carpet is an adroit production with classic green ground C Border

37-Beselin 127

Much rougher is a carpet once in the Wher Collection which was ultimately sold at the Davide Halevim sale for $117,090.Note the  miniature carpet in the top border.The Berlin rug has the typical border of the Small Medallion Ushaks,whereas the more archaic Wher/Halevim carpet has an SPH Quatrefoil,at which point Ushak and Konya(?)overlap.There are otherwise no classical Ushak carpets with a Kufi border.

38-Halevim Sale.Christies 14 February 2001(116)

To sum up briefly,the SPH carpets on a plain ground employ all borders from type A to H;those with a counterchange or checkerboard field use Borders A,B and D;the Lotto carpets use A to E.It may be that in the future the present scheme can be used to expose fakes.Stefano Ionescu revealed this in a Holbein carpet from Toronto


This niftily drawn item,first published in a Canadian rug catalogue in 1974,is actually an impossibility.There are no counterchange SPH carpets with the C or C1 border-the four known items are all on a plain ground.Presumably Mr Tuduc knew his Transylvanians very well!

 Part II