In 1878 Ramón Silvestre Verea García (1833-1899), a Spaniard, newspaper publisher in New York, patented a direct-multiplying machine, which proved to be the second patented machine of this type (after the machine of Edmund Barbour), 10 years before the first popular direct-multiplying machine of León Bollée.
Most early calculating machines carried out multiplication as a form of repeated addition. To multiply, say, by sixteen, one set the carriage at its rightmost position, turned the operating crank six times, shifted the carriage one position to the left, and turned the crank once. In direct-multiplying calculating machines, the operator had only to perform n operations when the multiplier was an n digit number.
Ramón Verea (see biography of Ramón Verea) is a very interesting person. Born in Spain, he lived for 10 years in Cuba and in 1865 settled in New York, working as a journalist in a magazine, agent for inventions and trading with Spanish gold and banknotes, that got him interested in calculation. Verea asserted that he did not make the machine to sell the patent or to put it to use, but simply to show that it could be done and that a Spaniard could invent as well as an American.
On September 10th, 1878, Verea received a U.S. patent №207918 for his machine. It seems, he manufactured also two prototypes, one of them sent together with the patent application to the US Patent Office, and second, which the same year (1878) was exposed and won a medal of the World Inventions Exhibition in Matanzas, Cuba. The newspaper Scientific American included an article about it. But then the sands closed over it. Verea never tried to market it. He just walked away and never invented anything else. As he said: "I just moved the desire to contribute something to the advancement of science and a little self-esteem. I am a journalist and not a scientist and also what I wanted to show... is already proven."
The prototype of the Verea's machine, sent to to the Patent Office
The prototype of Verea's machine, which was sent by the inventor to the US Patent Office, together with the application in July, 1878, was kept in the tanks of the headquarters of IBM in White Plains (New York) to be part of the collection begun in 1930 by the founder of IBM—Thomas Watson.
Verea's calculator was a made of iron and steel machine about 22 kilograms, 35 cm long, 30 cm wide and 20 cm high. It was able to add, multiply and divide numbers of nine figures, allowing up to six numbers in the multiplier and fifteen in the product. The multiplication solved through the direct method, based on a mechanism patented by Edmund D. Barbour in 1872. Verea saw how to do the whole multiplication in one stroke of a lever.
One of the patent drawings of the Verea's machine
The basis of his machine was a ten-sided metal cylinder (at the front side of the device you can see the two ten-sided brass cylinders that are mounted vertically). Each of the sides of the each cylinders has two columns of holes, with ten holes in a column. The holes come in ten sizes, with the largest and deepest representing 0, and the smallest and shallowest, 9. The holes represent multiples of a given digit.
Above the cylinders are two knobs that move in slots in the flat top of the machine. Pulling forward a knob rotates the cylinder below, so that the side facing the back of the machine has holes representing multiples of the digit desired. Behind this mechanism is a row of tapered pins. Pulling a lever at the back of the machine raises or lowers these pins in order to set the multiplier. Turning a crank on the right side moves the pins up to the faces of the cylinders and, where there are holes in the cylinder, allows the pins to enter to a certain depth.
Once the surface of a cylinder touched a pin, it pushed the pin, and a rack behind the pin, backward. Pins entering shallow holes reach the cylinder quickly and have a correspondingly greater effect on the rack. Pinions linked to the racks rotate correspondingly, rotating the result wheels at the back of the machine, with carrying occurring as required. Further turning of the crank restores the cylinder, racks, and pins to their original position.
During a demonstration, the device could solve 698,543,721 x 807,689 in twenty seconds, an amazing speed for the time.