LINSTEAN SOCIETY OF LONDON. 59 



though it must have been before 1792, when he was assassinated 

 by Aukarstrom. It cannot be maiutained that this is a mere copy 

 of the upper part of the Versailles' portrait, for the position of the 

 shoulders varies, the coat is of velvet, not silk, the necktie has 

 another form, and a spray of Linncea is placed in the button-hole, 

 instead of the hand, which does not appear in the smaller canvas. 

 In general the latter agrees with the Stockholm portrait : it seems 

 probable that the (iripsholm portrait was painted first, possibly as 

 a sketch ; it might have been done in Sweden, or after the painter's 

 return to Paris. 



It may therefore be assumed that the "Versailles portrait was 

 painted in Sweden, and is the original picture ; it was copied by 

 Pasch very soon afterwards, and disappeared from sight after 

 being exhibited at the Salon, till its reappearance fifty years ago. 

 It is possible that the Gripsholm portrait was painted also iu 

 Sweden, but the Stockholm picture must certaiidy have been 

 produced in Paris at a later period. 



As the Versailles portrait is practically unknown in this country, 

 tVie following details are taken from Prof. Tullberg's volume. It 

 was shown at the Salon in Paris in 1770, and came by purchase to 

 A'er.sailles before 18G1, and is numbered 4514. It has suffered 

 bad treatment, and was restored some years since. "When it was 

 being cleaned, it was discovered that the painting bore traces of 

 the cross of the Polar Star, which had previously escaped notice, 

 and since then it has been very unhappdy painted in, the crown 

 being omitted, and the cross placed flat on the ribbon, instead of 

 the true method of suspension. This explains why, in a series of 

 photographs issued in 1897, the decoration is wanting ; but it was 

 l)resent in 1906, as shown in a later photograph possessed by 

 Prof. Tuliberg. 



The account given by Dr. "W. Carruthers in our ' Proceedings,' 

 1905-6, pp. 67-68, set out the facts then available regarding the 

 Pasch copy, but more recent information has rendered that 

 account somewhat incorrect, as may be inferred from the foregoing 

 narrative. It may be restated thus : — 



Uno von Troil (1746-1803) was the son of an archbishop of 

 Upsala, who, alter taking his degree as Fhilosophue Maltster, went 

 on a foreign tour, during which he accompanied Sir Joseph Banks 

 and Dr. Solander to Iceland, iu 1770, and on his return to Sweden 

 published his ' Bref rcirande en resa till Island,' which aroused 

 great attention. He met Eoslin in Paris in 1771, and was there- 

 fore an acquaintance of the painter when he revisited his native 

 country in 1774-5. Von Troil became a court-chaplain in 1775, 

 and it seems practically certain that lie then induced Eoslin to 

 allow Pasch to make a copy of what we have termed the Versailles 

 portrait as a present to Sir Joseph Banks. It remained iu Banks's 

 possession till his death in 1820, when it passeil to Eobert Brown, 

 under the proviso of Banks's will, that the household furniture in 

 the Soho Sqtiare establishment should pass to Brown, upon Lady 

 Banks ceasing to reside there after her husband's death. Brown 



