_Third Exception._--In cases where there is no practical utility in comparing one very large number with another, as in the case of the distances of the planets from the sun, mere round numbers may suffice, yet astronomers must know such numbers with exactness. But in regard to all mundane affairs, the pupil must throw off the character of scholar and assume the license of children, if he attempts to express large numbers, as of populations, &c., by "guessing," or, what is the same thing, by only giving round numbers. The Brooklyn Suspension Bridge is 5989 feet long, and the Forth Bridge, which crosses the Firth of Forth in Scotland, is 8296 feet long. Now, instead of saying that the former is _about_ 5000 feet long, why not say 5989 feet long? [(5) {L}ong (9) {B}ridge (8) O{f} (9) {B}rooklyn.] And instead of saying that the latter is _about or somewhere in the neighbourhood_ of 8000 feet long, why not be exact and say 8296 feet long? [(8) {F}orth's (2) {N}ew (9) {B}ridge (6) {Sh}own. It was completed in 1890.] No one who has not had experience in dealing with thousands of poor memories, as I have had, can realise the fact that in most cases of poor memories _the facts themselves are often possessed_, but are mostly _unrecallable_ when wanted. I have tried to teach pupils how to find analytic date or number words _without any previous training in In., Ex., and Con._, and 99 of all such attempts have always been failures. The 100th case, which succeeded, only confirmed the rule. On the other hand, I have always found that these failures become successes after a thorough practical training in In., Ex., and Con., such as I have already given. In fact, I never had a pupil who became proficient in the use of In., Ex., and Con., who did not arrive at the use of analytic number words without any specific directions from me. But I think, on the whole, that it is the better way to _combine_ direct and specific training in analytic number words, with a previous exhaustive general drill in In., Ex., and Con. The rules hereafter given must be carefully studied and every example painstakingly examined. After studying my formulas let the pupil endeavour in _each case_ to find a better one himself. If the pupil acts on my advice, he will know how to be always _sure_ to think of the needful related or including facts for finding analytic date words, phrases, or sentences. The different processes for dealing with dates or numbers may be classified as follows:--