Early in the spring of 1894, New York State lawmakers set out to investigate New York City’s corrupt police force. Rockland County senator Clarence Lexow headed the committee, which spent the year hearing stories about seedy dealings in the city. The investigators found expected forms of graft—brothel keepers and saloon owners cutting the police a share of their profits, for instance—but also got a glimpse of smaller street-level payoffs, previously not much on their radar. Some of these were simple, made to avoid the city’s new food permits and licenses. But much stranger was what investigators learned from witnesses like Morry Oswitz, an erstwhile stained-glass-window salesman with bad lungs, who told the committee about his days spent hidden in the back room of a candy store on the Lower East Side, managing a popular form of community-based gambling often called policy.

“Policy” was a common way to refer to the numbers game, a system that involved bettors “insuring” a set of numbers by putting a few cents down on the chance that these numbers would be drawn in one of the country’s legal lotteries. Oswitz was a policy writer, which meant he sat in that back room awaiting bettors to bring in their selections and money. Writers often also dealt with men employed by shops as runners, who collected bets from around the neighborhood. A writer would record the picks and, if any of them hit, manage the payout. The idea traced back at least to eighteenth-century Londoners who were too poor to buy a whole lottery ticket but could afford to take out a small side bet on a few numbers.

The game’s reach grew in America in the nineteenth century, and it had become vast in New York by the 1890s. By then, a bettor could take out “insurance” on numbers from a variety of sources. Some shops pulled numbers from Kentucky lottery cards. Others paid out wins when bettors hit numbers that appeared in racetrack receipt totals. Players wagered so many pennies and nickels—and winners were so few—that the game turned healthy profits for the seven-hundred-odd policy shops in the city, as well as for the beat cops who allowed them to exist. One witness told the committee that on Seventh Avenue you could find a policy shop on every block. “Did you ever meet any of your friends, the police, in those policy shops?” he was asked. “Oh yes,” he responded, “many times.”

A policy writer named J. Lawrence Carney described in his testimony how widespread the game had become. Women played, he said, and children—even a handful of “prosperous people.” Policy play spread through city neighborhoods. “Downtown it is the poor class, the Jewish people,” Carney explained, “uptown it is the Negroes, and in Little Italy they are Italians.” Poor neighborhoods were the best: “They do not play such big amounts, but the quantity more than covers that.” The bets were small, the payout relatively high: many were willing to give up a nickel in hopes of the ten dollars they might see if by some grand stroke of luck their chosen numbers should hit.

As hearings dragged on, the committee’s curiosity grew regarding how bettors chose the numbers they played. There seemed to be some kind of system to policy play, even if it wasn’t rational or mathematical. Witnesses explained that bettors were playing specific, named combinations that would be commonly understood at a policy shop. The names of the types of combinations were suggestive of horse racing: when players bet a combination of two numbers, it was called a saddle; when they bet four numbers, it was called a horse. But the most common play was a combination of three numbers, called a gig. “There are thousands of gigs,” Oswitz told the committee, and they all were given names. “The policy writer names it, and if it hasn’t any name, he gives him a name.” He revealed during questioning a few he had seen as a policy writer: the beer gig (which indicated a bettor playing the numbers 4-16-21); the policeman’s gig (28-35-67); the sick gig (10-20-30); the dead gig (9-19-29); and the working gig (14-26-42). There were gigs named after the days of the week, for horses and dogs, after names of men and women, for a Brooklyn saloon­keeper, and for the Sing Sing cell number of a fallen crime boss.

Investigators sought to understand the curious logic that tied particular numbers—seemingly arbitrarily—to stray language, imagery, and experiences of daily life. Oswitz offered to “count the gigs off,” but a police board member stopped him short. “I don’t want you to give me the numbers of the gigs,” he said. “I want you to give me the names.” That disenfranchised folks might be attracted to low-cost gambling seemed clear enough. But the association of particular numbers with names and meanings seemed to abandon rationality for accident and chance. “If a policeman ran up against the Broadway cable car and got the worst of the collision,” another investigator asked, “you mean to say the people would run in and play the police gig?” Carney confirmed. “They would take the police gig, take the number of the car, and run the combination.”

Oswitz revealed a further, pervasive method, one derived from subconscious life. “Everybody has a dream,” he said, “and next morning they come in and tell them to the policy writer, and the policy writer gives them a gig for it.” Carney explained how this worked. “There are dream books, Common Sally and Three Witches and Wheel of Fortune. Now, those books have every word in the dictionary, I guess, and they will have the lucky number opposite; if a fellow dreams he has seen a horse, if he is riding horseback, he will pick up the book, find forty-nine, and take out that.”

Lexow’s investigators had stumbled on a contemporary version of the old art of oneiromancy, or divination through dreams. Their attempt to uncover one secret world, of urban corruption, had revealed another: an alternative logic of chance and luck operating just beneath the measured rhythms of modern life. But it was tough for a legal-minded state senator to compute. One investigator soon wanted to know if there was a “Lexow gig.” (There was not.) Investigators’ questions revealed deep incredulity about dream books—how could anything translate a dream of horses into a bet at a backroom policy shop?

They were not the first to wonder.

A half-century earlier, in 1851, Jonathan Harrington Green, an opportunistic cardsharp turned spurious antigambling reformer, published a Report on Gambling in New York. Green mapped some six thousand gambling houses in the city; by his calculation, those houses each year spelled the ruin of at least thirty thousand young men who fell into gambling’s “mad whirlpool of passion and temptation, fatal alike to their temporal and eternal interests.” Green used his own past experiences of fleecing gambling neophytes to make a case that associated luck with suckers and fraud. His campaign also made clear that part of the problem was a belief in dreams, and in his report he paused to describe a manual that had become popular in America by the time of his writing, Old Aunt Dinah’s Policy Dream Book: Comprising a Brief Collection of Dreams, Which Have Been Interpreted and Played with Wonderful Success to the Dreamer.

Published manuals of dream interpretation were an old story. One well known from antiquity was by a second-century physician from Ephesus named Artemidorus, who gathered interpretations into his five-volume Oneirocritica. Old Aunt Dinah trafficked in this ancient art but offered modern twists. The character of Old Aunt Dinah had been created by the book’s publisher with the intention of lending a kind of blackface minstrel expertise to interpretations. But most important for Green, Dinah’s dream interpretations weren’t just about foretelling the future; the book translated dreams into specific numbers to play at the policy shop.

Green quoted at length. “To dream of seeing a person’s corpse lying on a bed is a good dream for the dreamer. If the scene is viewed with surprise and awe, it is good to play fifty-sixth fourth, or your age and the person’s age in a gig.” The book was very specific. “To dream of seeing a Negro man, or one of very dark complexion, is a favorable token for the dreamer for fourteen first,” it said. “If the person be very dark, inclining to jet-black, it is more favorable. If the viewing of the person is exciting to you, the number may with safety be anticipated very soon, and would advise the player to play heavy, as it is a sure thing.”

Old Aunt Dinah knew to give her readers happy news. The book was full of good dreams and sure things. Another entry interpreted “sleeping with a young lady or gentleman, which would not be proper to do in reality.” This dream “is very certain for fifty-one first, or the age of the person you dream you sleep with and your own age last. This is a very good dream.”

The book’s success brought flocks of imitators, all ready to offer easy predictions on money, love, and the weather. Dream interpretation was everywhere. Printers added chapters to almanacs; salesmen tossed pamphlets in with patent medicines; and publishers added little books to catalogues of joke books, popular recitations, and scripts for minstrel shows.

American dream books from these nineteenth-century boom years were not particularly original. Some writers gave a nod to Artemidorus. Others looked locally to astrologers, weather prophets, or various American versions of The New Book of Knowledge, a popular manual on dreams first produced in Boston in the late eighteenth century. Many simply reprinted interpretations pulled from seventeenth-century chapbooks.

These texts seem inconsequential, perhaps, but dream books became steady-selling money­makers for popular printers in America as well as across the Atlantic in England. Scottish journalist Charles Mackay included dream books among the follies he described in 1852 in his Memoirs of Extraordinary Popular Delusions and the Madness of Crowds. “Two books in England enjoy an extraordinary popularity and have run through upward of fifty editions in as many years in London alone, besides being reprinted in Manchester, Edinburgh, Glasgow, and Dublin,” Mackay wrote. “One is Mother Bridget’s Dream-Book and Oracle of Fate; the other is the Norwood Gipsy,” another dream book. He marveled at the “demand for these works, which are sold at sums varying from a penny to sixpence, chiefly to servant girls and imperfectly educated people, all over the country.” The figures astounded him: more than 11,000 sold annually on average over the past thirty years, more than 330,000 total.

Mackay scoffed at servant girls discussing dreams over morning tea. But he paused nonetheless to decipher oneiromancy’s logic. “The rules of the art, if any existed in ancient times, are not known,” he wrote. “But in our day, one simple rule opens the whole secret. Dreams, say all the wiseacres in Christendom, are to be interpreted by contraries. Thus, if you dream of filth, you will acquire something valuable; if you dream of the dead, you will hear news of the living; if you dream of gold and silver, you run a risk of being without either; and if you dream you have many friends, you will be persecuted by many enemies.”

This interpretation by inverse appeared throughout American dream books in the 1860s. Among the most elaborate was The Golden Wheel Dream-Book and Fortune Teller, by the happily named Felix Fontaine. “To dream of making a sudden fortune,” wrote Fontaine, “is a sign of want. All dreams of this kind go by the rule of contrary.” To dream of an enemy fortells success (in proportion to the injury sustained in the dream); of a contagious disease brings luck; and of jail guarantees a future of honor, since the “rule of contrary” governs all our dreaming selves.

The rule of contrary reveals a broader function of dream books as a genre, one that helps explain why they found such ready audiences among the servant girls in England as well as the “poor class” people hunting for winning numbers in New York City. If what appeared in a dream meant the opposite, then the logic of dream books allowed poor, hopeful readers to imagine a world turned upside down. A sudden fortune was said to mean a sign of want; perhaps a life of want indicated fortune just around the corner.

The invented experts of this topsy-turvy dreamworld were another important piece of the puzzle. Old Aunt Dinah was a typical kind of exotic seer used by printers to give credence to the mystical art being wielded; dream-interpreting witches, gypsies, and seers of other popular books included Uriah Konje, Prince Ali, Moses Magical, Harlem Pete, Policy Pete, Policy Joe, Old Moore, the Mystic Oracle, and Madame Fu Futtam. They were there to give an oracular air to simple pronouncements and to gesture at forces operating in the cracks of a world that otherwise suited clerks and bureaucrats. They were figures that could see outside the daily demands of bankers, bosses, and landlords.

As policy play expanded through the nineteenth century, so did dream books like Aunt Dinah’s that equated oneiromancy with numbers to bet. The rise was buoyed by an upswing in cheap printing and by swelling numbers of poor New Yorkers—immigrants from Ireland and Italy, and African Americans barred from many of the city’s more lucrative occupations. Dream books hit a peak of popularity about the time Lexow and his investigators heard about them in their 1894 testimony. By then, many American dream books had dispensed with narrative interpretation altogether, simply reducing the form to lists of words associated with numbers to bet. A large part of Felix Fontaine’s book became just a “List of Dreams Without Interpretations but with the Numbers They Signify”: anchovies, 73, 1, 62; bathroom, 11, 76, 1; dung-forks, 25, 62; history, 49, 58; helping anyone, 6, 73.

The books followed their markets well into the twentieth century. By 1940 writer Claude McKay noted that “Harlem is haunted by numbers. Dreaming of numbers is an inevitable condition of the blissful state of sleeping,” and policy play was “the most flourishing clandestine industry in Harlem.” Makers of magic oils and voodoo creams added dream books to their product lines and made room on their dream lists for the things that had entered dreamers’ lives: numbers for typewriters, elevators, adolescents, airplanes, library cards, motor­cycle cops, and the Ku Klux Klan. As world war took over lives both awake and asleep, books gave dreamers numbers for U-boats, navy yards, and disabled soldiers.

I​t’s easy to dismiss dream books as trivial junk washed up in the wastes of popular culture. Yet their persistence catches a trace of the luck we have sometimes chased out of our calculations of rational marketplace success. Late in the dark year of 1942, as dream books were adding additional war lingo to the psychic lives of bettors, writer E.B. White was finishing up five years spent on a farm in Maine, from which he had been filing a monthly column for Harper’s Magazine called One Man’s Meat. “Luck is not something you can mention in the presence of self-made men,” he wrote, a line that would later become a favorite of quote books. Nor is luck “something that you can talk to economists about, because it does not fit into the hard-earned grooves of social thought.”

Easier to find luck in the mushy minds of dreamers. Around the same time, entomologist and antiquarian Harry B. Weiss was looking over dream books at the New York Public Library. He thumbed through a collection of fifty-seven dream books, many similar to those that had troubled Charles Mackay. He saw “no rules and no system of interpretation”; even their prefaces were “masterpieces of nothingness.” But Weiss was gentler than Mackay and shrugged off dream books as harmless folk beliefs. “All of us acquire folklore of one sort or another on such things as capitalism, cigarettes, political platforms,” he concluded, “even when our beliefs are shown to be without foundation. A serious belief in dream fortune-telling is simply part of the behavior pattern of some people.”

But in other contexts, dream books seem far more charged than mere “nothingness,” not to be dismissed as weightless habit. In his great midcentury novel, Invisible Man, Ralph Ellison locates dream books back on the streets of Harlem. The novel’s narrator comes upon a “sullen-faced crowd” watching as two white men evict an elderly couple from their apartment. The narrator provides an inventory of the “clutter of household objects” piled on the sidewalk—a material outline of the history of African America. Knocking bones from a minstrel show, a straightening comb, an Ethiopian flag, a tintype of Lincoln, a plate from the St. Louis World’s Fair, a bent Masonic emblem, three voided insurance policies, a newspaper portrait of Marcus Garvey, a Bible, a man’s free papers. Also there: “a dime pierced with a nail hole so as to be worn about the ankle on a string for luck,” a rabbit’s foot, and a dream book.

Spurred by the angry crowd, “the invisible man” becomes a street speaker. “Yes, these old folks had a dream book,” he says,

but the pages went blank and it failed to give them the number. It was called the Seeing Eye, The Great Constitutional Dream Book, The Secrets of Africa, The Wisdom of Egypt—but the eye was blind, it lost its luster. It’s all cataracted like a cross-eyed carpenter and it doesn’t saw straight. All we have is the Bible and this Law here rules that out. So where do we go?

Read by these remnants, the old couple’s lives have time and again seen glimmers of opportunity offered by the rational path of working life—emancipation, day jobs, insurance policies—come to nothing. They have gathered all the tools they can access to get ahead in an economy with odds stacked heavily against them, but now, kicked to the curb in their late years, the dream book seems as compelling a guide toward some hope of return as any the old couple was likely to find. Though it has failed them too.

“The Society of Movers and Doers is a very pompous society indeed,” E.B. White wrote in his essay, “whose members solemnly accept all the responsibility for their own eminence and success.” Such a society might hold the old couple fully responsible for their failures to make something out of themselves. Fooled by rabbits’ feet and coin amulets—and worse, a dream book, in whose numbers they should have known better than to put their faith—they are suckers who placed too much stock in the fraud game of luck.

But the sullen-faced crowd of Invisible Man doesn’t leave the old couple behind to pay for the mistake of their faith. Instead, the crowd forms a mob on their behalf. “We believe in brotherhood,” calls out one man who joins in the fray.

Since the 1940s, economists have opened up their hard-earned grooves to research what White called “the wine of luck.” Findings indicate that those who see luck as an element of their good fortune are more generous, while some successful people who believe themselves to be self-made are more selfish. Perhaps this reflects a broad kind of rule of contrary functioning both in and out of dreams. For even if playing a gig out of Old Aunt Dinah is no sure thing, we could do worse than listen to Felix Fontaine’s generous interpretation of giving alms: “To dream they are begged of you and you refuse them shows want and misery,” he wrote. “But to dream you give them freely is a sign of joy and long life.”