Our use of the numerals which we call "Arabic" is comparatively recent, and it is believed that the Arabs got these numbers from India several centuries after the Koran was written, or about eight hundred years after Christ. But the fact that the Greeks and others used the letters of their alphabets for numerals, caused the order in which they were written to remain fixed. If _alpha_ stood for 1, _beta_ for 2, _gamma_ for 3, _delta_ for 4, and so on up to ten, then a newly coined or newly adopted letter could not be inserted without great confusion; it had to be tacked on to the end of the alphabet. So, when scholars find in inscriptions letters, adopted from another alphabet, which stand out of their natural order, they can make a shrewd guess at the century in which the inscription was made. Suppose an alphabet, which is also used for numerals, loses a letter in the course of time, because there is very little or no use for it; then that letter is still of service for a numeral, and it can not be dropped as a number, though it drops out as a letter. When it is found still employed as a numeral, it reveals some of the history of the alphabet to which it once belonged. These are only a few of many methods of determining the age of a given inscription. Old coins are very useful in settling what the alphabets of various nations were at different epochs. Our own numerals are extremely convenient for ordinary arithmetic. Algebra, in which letters stand for numbers, is useful for abstract reasoning in mathematics; it treats of the properties of numbers in general. Whether the Indian numerals were originally part of some ancient alphabet, or a series of shortened signs originally somewhat like the Roman numerals that we still use, is not really decided.