In the second place these invasions were made by peoples who, though not literary themselves, had a somewhat high regard for literature, and there could have been no such reason for destroying histories as might exist where one dynasty was forcibly ejected by another hostile one. In the third place the monasteries were the great literary centers during the centuries preceding the rise of the present dynasty, and we may well believe that the Mongols would not seriously molest these sacred repositories. On the whole then we may conclude that from the year 57 B.C. Korean histories are fairly accurate. Whatever comes before that is largely traditional and therefore more or less apocryphal. One of the greatest difficulties encountered is the selection of a system of romanisation which shall steer a middle course between the Scilla of extreme accuracy and the Charybdis of extreme simplicity. I have adopted the rule of spelling all proper names in a purely phonetic way without reference to the way they are spelled in native Korean. In this way alone can the reader arrive at anything like the actual pronunciation as found in Korea. The simple vowels have their continental sounds: _a_ as in “father,” _i_ as in “ravine,” _o_ as in “rope” and _u_ as in “rule.” The vowel _e_ is used only with the grave accent and is pronounced as in the French “_recit_.” When a vowel has the short mark over it, it is to be given the flat sound: _ă_ as in “fat,” _ŏ_ as in “hot,” _ŭ_ as in “nut.” The umlaut _ö_ is used but it has a slightly more open sound than in German. It is the “unrounded o” where the vowel is pronounced without protruding the lips. The pure Korean sound represented by _oé_ is a pure diphthong and is pronounced by letting the lips assume the position of pronouncing _o_ while the tongue is thrown forward as if to pronounce the short _e_ in “met.” Eu is nearly the French eu but with a slightly more open sound. As for consonants they have their usual sounds, but when the surds _k_, _p_ or _t_ in the body of a word are immediately preceded by an open syllable or a syllable ending with a sonant, they change to their corresponding sonants: _k_ to _g_, _p_ to _b_ and _t_ to _d_. For instance, in the word _Pak-tu_, the _t_ of the _tu_ would be _d_ if the first syllable were open. No word begins with the sonants _g_, _b_ or _d_. In Korean we have the long and short quantity in vowels. _Han_ may be pronounced either simply _han_ or longer _haan_, but the distinction is not of enough importance to compensate for encumbering the system with additional diacritical marks. In writing proper names I have adopted the plan most in use by sinologues. The patronymic stands alone and is followed by the two given names with a hyphen between them. All geographical names have hyphens between the syllables. To run the name all together would often lead to serious difficulty, for who would know, for instance, whether _Songak_ were pronounced _Son-gak_ or _Song-ak_?