It was allegedly painted in 1638 when Captain John Underhill made a return visit to London.
In the version of the portrait published in 1932 the words "John Underhill of Warwickshire" appear in the upper right portion, but these are not found on the portrait itself. On the back of the portrait was a label (reproduced on the reverse of the frontispiece to Volume 1) which reads as follows:
"A Portrait of John Underhill Esq. of Warwickshire son of Thomas of Barton on the Heath and brother of Sir Edward Underhill Kt of Eatington. He, John, went to New England with Governor Winthrop to train the people in military discipline and was appointed Captain in command of the New England detachment by Sir Henry Vane. He was made Governor of Dover in Piscataqua. He died at Killingworth 1672."
A second label reads "Relined and restored October 1866."
The unexpected appearance of this portrait was met with surprise and also with some skepticism since there was no provenance to show where the picture had been for nearly three hundred years, and no known Underhill will, inventory, or other family document had ever mentioned its existence. Among those who were skeptical was the scholar John H. Morrison, whom Mr. Taylor had commissioned to research the Underhill ancestry in England. In his 1932 book, The Underhills of Warwickshire Morrison makes no mention of any portrait of the Captain.
In spite of this Mr. Taylor had the portrait published, and it hung in his house at Locust Valley until his death in 1960, after which it became the property of the Society for the Preservation of Long Island Antiquities (SPLIA), which still owns it and displayed it for many years.
When I first began to research my Underhill ancestry in the 1950s, I was naturally much interested in the portrait, as relatively few ancestral likenesses survive from colonial times. I noticed, however, that the label identifying the portrait was in a much later handwriting than that of the 1630s. The names of the Captain’s father and brother given in the label are incorrect but match the names given in the Captain’s biography published in 1882 in the British Dictionary of National Biography (the correct ancestry remained unknown until the research of Mr. Morrison). It, therefore, appeared that the label might be younger than the painting by at least 274 years.
The mystery deepened around 1966 when my mother saw a color reproduction of the portrait on display in the gift shop of Burdine’s, a Miami department store. There was a label on the back in a language which she couldn’t read, but it clearly did not mention the name Underhill. On my next trip to Florida, I went to the store and found that the label, in Italian, stated that this was a copy of Van Dyke’s portrait of Viscount Stafford, from the original now in the Museum of Art, São Paulo, Brazil.
At the New York Public Library I read the Dictionary of National Biography’s account of Stafford, which identified him as William Henry Howard (1614-1680), a member of the well-known Howard family whose senior member is the Duke of Norfolk. The account revealed that he was twice painted by Van Dyke, one portrait then (1882) being owned by the Marquis of Bute and the other by the Duke of Norfolk. It is the latter portrait that we are concerned with. It was reproduced in 1908 in Brenan and Statham’s The House of Howard, but as an engraving in which the features are slightly distorted so that I was still uncertain, it was the same portrait.
About 1973 I noticed an entry in the catalog of a New York print shop for a copy of Van Dyke’s Viscount Stafford. This turned out to be a full-color reproduction (printed in Italy from the original at São Paulo), and it matched the alleged Underhill portrait exactly. There was no longer any question that the alleged portrait of Captain John Underhill was a copy of Van Dyke’s portrait of Viscount Stafford. I purchased a copy of this print and it is now in the collection of the Society.
About 1980 a representative of SPLIA advised me that tests on the portrait in their possession had shown that the paint was about 100 years old, not 300. Tests on other alleged 16th and 17th century Underhill portraits given to SPLIA from the Taylor estate showed similar results (for these portraits see the Bulletin,