The plaintiff in Robinson, also known as the Howland Will Case, was Henrietta “Hetty” Robinson (later Green) of New Bedford, Massachusetts. Robinson’s family was quite wealthy, having made its fortune in the whaling business. When her father died in 1865, she inherited nearly a million dollars plus the right to income derived from trusts worth several million more. Her investment savvy and extreme thrift caused her to be dubbed both America’s richest woman and its greatest miser during the Gilded Age. Some called her “the Witch of Wall Street.”

Shortly after her father’s death, Robinson’s aunt, Sylvia Ann Howland, passed away as well. Howland, with whom Robinson was very close, left an estate worth approximately two million dollars. Her will placed about half of the money in trust for the benefit of Robinson and made various bequests of the remainder. Robinson challenged the will in probate court on the ground that her aunt had been incompetent to make it, but soon withdrew her appeal.

A short time later, Robinson filed an equity suit in the U.S. Circuit Court for the District of Massachusetts, seeking specific performance of an agreement she claimed to have made with her aunt. Under the agreement, each woman would bequeath all her property to the other, they would maintain copies of each other’s wills, and neither would make another will without informing the other. In support of her claim, Robinson produced a will, alleged to be Howland’s, that predated the will admitted to probate by several years. The earlier will left the entire estate to Robinson. An addendum, handwritten by Robinson but bearing Howland’s signature, stated that the will revoked all others, whether made before or after. The addendum explained that Howland feared her caretakers would take advantage of her weakened state by forcing her to make another will. As it was her “lifetime wish for [Robinson] to have my property,” she wanted her niece to keep the will in case it became necessary to challenge a later one.

The executor of Howland’s estate put forth several arguments in defense, one of which was that Howland’s signature on the addendum was a forgery. The defense called a host of expert witnesses—including chemists, engravers, accountants, bankers and handwriting experts—to testify that the signature on the addendum was perfectly identical to the signature on the will (which was conceded to be genuine) and that such an exact reproduction was not possible. All the experts agreed that the signature in question must have been traced from the original.

The most unusual expert witness was world-renowned mathematician Benjamin Pierce, who taught at Harvard University. Pierce declared with “the utmost degree of confidence” that the signature was a fake. He used statistical evidence to support his conclusion, testifying that the phenomenon of an exact reproduction of a signature would occur by chance only once in “two thousand six hundred and sixty-six millions of millions of millions of times.” “This number far transcends human experience,” he continued. “So vast an improbability is practically an impossibility.” Pierce added that having both signatures placed at the exact same distance from the edge of the page increased the improbability “probably a hundred-fold.”

Despite the testimony of these expert witnesses, and a great many on the plaintiff’s behalf, the jury never had the opportunity to deliberate on the evidence and reach a verdict on the forgery issue. Justice Clifford dismissed the case on other grounds, citing a Massachusetts state law barring a person who made a contract with someone now deceased from testifying in their own behalf in a suit against an executor. Without Robinson’s testimony, Clifford ruled, the evidence she presented was insufficient to sustain her allegation of an agreement with her aunt.

Robinson v. Mandell is notable primarily because it is believed to be the first federal court case in which statistical evidence was presented. In contrast with the present day, no body of jurisprudence concerning the use of scientific evidence in court yet existed. As Clifford acknowledged with respect to the expert testimony, “[s]ome of the questions discussed were new, and it must be admitted that they are highly important as affecting the rules of evidence in cases where the genuineness of written instruments is in contestation,” but because the court did not have to determine the veracity of the signature, “the court is not inclined to decide those questions in this case.”