The Cigarette
A Richmond firm offered seventy-five thousand dollars for a machine to roll cigarettes, then would not use the one it got, and a rival in Durham built an empire on the lease.

In 1876 the Richmond firm of Allen & Ginter announced a prize of seventy-five thousand dollars for the first machine that could roll and cut a complete cigarette. The money was serious, and behind it sat an arithmetic the trade could not get around. Cigarettes were made one at a time by hand, by rollers skilled enough that the Virginia and Carolina factories recruited immigrants from eastern Europe to do the work, and a good one produced about four a minute. Demand was climbing faster than fingers could move.
A teenager in Roanoke County read the announcement and could not put it down. James Albert Bonsack had grown up around the machinery of his father’s woolen mill on the Salem road, entered Roanoke College in 1878, and left after a year to set up a workshop in the mill instead. The mill is where the answer came from. Its carding machine, built to comb raw wool into an even web, solved the problem that had defeated everyone else, which was how to lay down tobacco in a consistent quantity. He filed on 4 September 1880, a month short of twenty-one, and the patent issued the following March as number 238,640.
Bonsack did not build a faster pair of hands, and the distinction is the whole invention. A hand roller finishes one cigarette and starts another. Bonsack made the cigarette continuous and cut it afterwards. Shredded tobacco went onto a belt ruled into equal divisions so that equal weights were laid down along it, then passed between rollers clothed in card-cloth that combed it out, and dropped from a carrier that slid back and forth as it worked so that any clot was scattered along the belt below instead of landing in one place. All of that exists to serve one requirement. The rod has to be the same thickness everywhere, because a cigarette with a thin place burns fast and a cigarette with a thick place will not draw.
The evened tobacco fell into a trough where a second belt gathered it, curled around it, and drew it through a tapering tube until it emerged compressed into what the patent calls a rope. Paper came up beneath it from a spool and entered a second tube alongside, where a spiral groove wrapped it around the rod; at the moment the wrap closed, the tube opened a narrow lip so a rubber wheel could lay paste along the underneath edge, then shut again to press the lap down. Last came the knife, a disc on a crank geared to swing past the emerging rod far faster than it travelled the rest of its circle, because a rod that meets a slow blade buckles and bursts.

The specification states the result without a single adjective. In practical operation, it says, the machine has shown a capacity of a hundred thousand cigarettes in a working day of ten hours, which is something over a hundred and sixty a minute, or what Duke Homestead reckons the labour of forty-eight hand rollers. The patent office still files cigarette machinery under a class called machines of the continuous-rod type, and every machine-made cigarette since 1881 has been a length of that rod cut to size.
The object of my invention is to provide a machine which shall uniformly feed and distribute the tobacco upon a continuous paper ribbon, then form the same into a continuous roll, then paste the paper around it, and, finally, cut off the same into definite lengths, all in a series of consecutive operations. This general result has heretofore been attempted, but so far as I know with but little success.
My invention is designed to provide such a construction of machine as will successfully accomplish the above result, and I have produced such a machine, which, in practical operation, has exhibited a capacity of one hundred thousand cigarettes per day of ten hours.
James Albert Bonsack, US Patent 238,640, “Cigarette-Machine.”
Allen & Ginter did not use what they had paid for. A machine was tested in their Richmond factory in June 1881; they ran it, judged it a failure, and went on selling the dearer hand-rolled article on the reasoning that smokers would not take a cigarette no hand had touched. They were still refusing to fit out the factory with it in the middle of the decade, by which time the argument was over. The firm that took it instead was in Durham, and had entered the cigarette trade in 1881 for a defensive reason: W. T. Blackwell & Company, a few streets away, was selling Bull Durham smoking tobacco faster than W. Duke, Sons and Company could sell anything at all.

In 1884 the Dukes leased two of the rejected machines. They jammed, so the Bonsack Machine Company sent a mechanic named William Thomas O’Brien down to Durham, where he worked the trouble out of them alongside James Buchanan Duke, halved the cost of making a cigarette, and eventually had the Durham factory turning out as many as four million a day. Then Duke made the move that actually decided the industry, and it was not a mechanical one. He committed his entire output to Bonsack machines in return for a royalty below the rate the Bonsack company charged anybody else. A lower cost per thousand let him undercut his rivals and still advertise at a level one journalist of the day called “a proportion of his profits which seemed appalling to more conservative manufacturers.”

By the late 1880s five firms were making ninety per cent of the country’s cigarettes, and all five were bleeding money into advertising to hold their share. In 1890 they folded into one another as the American Tobacco Company, capitalised at twenty-five million dollars, with Duke as its president, and within fifteen years it had taken most of the plug, snuff and smoking-tobacco trades as well. In 1911 the Supreme Court found the combination in violation of the Sherman Antitrust Act, and that November a New York circuit court divided the assets among a reduced American Tobacco, Liggett & Myers, P. Lorillard and R. J. Reynolds. The remedy did not calm the market, since four companies now had to fight for it, and Reynolds answered in 1913 with Camel.
American adults smoked a few dozen cigarettes a head in 1900. Consumption climbed through two world wars and peaked in the early 1960s at roughly 4,400 a year for every adult in the country, by the Surgeon General’s reckoning. Two studies published in 1950, Wynder and Graham in May and Doll and Hill in September, were large enough and clean enough to make the association with lung cancer hard to argue with, and in January 1964 the Surgeon General’s advisory committee reported. Per capita consumption has fallen in most years since, to about 1,200 by 2012. The appetite was older than the machine. What the machine took away was cost and scarcity as limits on it, which is why the twentieth century inherited the cigarette as a mass article rather than a minor one.