The development of the stereotyping process, however, brought to light difficulties with type. Foundry type was sold with the shoulder beveled off for ordinary printing and this was not favorable for stereotyping. The type founders would not make the high spaces and quads which were needed. As the best way of meeting these difficulties, Bruce, in 1815, went into company with the Starr brothers for the manufacture of type. It soon appeared that type founding and stereotyping promised to be more profitable than printing and Bruce sold out his printing establishment to devote himself to the other branches. Bruce and the Starrs were unable to agree and the partnership was soon dissolved, Bruce deciding to carry on the business alone in spite of the difficulties of every sort which surrounded it. In addition to the business conditions of the time, neither of the Bruces had any practical knowledge of type founding and the matrices of their only complete font were stolen, presumably by someone who was interested to secure their failure. It needed more than that, however, to discourage this persistent Scotchman. George Bruce set himself at work to learn punch cutting and mould making. His first efforts were crude, but he had an artistic temperament, a critical spirit, and a practical knowledge of printers’ needs. By these qualities and his own persistence he soon became very proficient. By 1820 the Bruce foundry was the best in the country, doing better work than even Binney & Ronaldson at that time. In 1822 Bruce undertook to remedy the confusion in sizes which was then and for a good many years a source of difficulty, annoyance and expense to printers. He devised a scientifically correct system by which the size doubled with every seven sizes of the system. This was uniform throughout, so that wherever you touched the system, you found any given type twice as large as the seventh below and half as large as the seventh above. In spite of the fact of the simplicity and scientific correctness of the system it did not prove suited to commercial work and was not adopted. In 1828 William M. Johnson had invented a type-casting machine in which a pump forced the liquid metal into the mould, giving the type a sharper face than was possible with hand casting. The machine was a step in the right direction, but was crude and imperfect. White took it up and tried to improve it, but he did not succeed in removing its fundamental defects. The types were not cast solid. Being hollow they were light and too weak to withstand the pressure of the presses. The first successful type-casting machine was made by David Bruce, Jr., in 1838, in development of the Johnson idea. George Bruce bought David Bruce’s patents and used the machine until 1845, when David Bruce made further improvements and produced the type of machine which is now in general use not only in this country but in Europe, where the method was soon adopted.