But the _Romans_ had both sounds, the _flat_ as well as the _sharp_, g as well as k. How did they express them? Up to the second Punic War they made the rounded form of the Greek Γ, out of which the letter C has arisen, do double work, and signify k and g equally, just as in the present English th is sounded as the Greek θ,[45] and as dh;[46] in proof whereof we have in the Duillian column, MACESTRATOS = MAGISTRATOS, and CARTHACINIENSES = CARTHAGINIENSES. Thus much concerning the power and places of the Latin c, as opposed to the Greek γ. But this is not all. The use of _gamma_, with the power of k, made _kappa_ superfluous, and accounts for its ejection in the _Etruscan_ alphabet; a fact already noticed. Furthermore, an addition to the Etruscan alphabet was required by the existence of the sound of g, in Latin, as soon as the inconvenience of using c with a double power became manifest. What took place then? Even this. The third letter was modified in form, or became a new letter, c being altered into g; and the new letter took its place in the alphabet. Where was this? As the _seventh_ letter between f (_digamma_) and h (_hæta_). Why? Because it was there where there was a vacancy, and where it replaced the Greek _zæta_, or the Hebrew _zayn_, a letter which, _at that time_, was not wanted in Latin. d. _Change of order._--As far as the letters c and g are concerned, this has been explained; and it has been shown that change of order and change of power are sometimes very closely connected. All that now need be added is, that those letters which were _last_ introduced from the Greek into the Roman alphabet, were placed at the end. This is why u, v, w, and y come after t--the last letter of the original Phœnician, and also of the _older_ Greek. This, too, is the reason for z coming last of all. It was restored for the purpose of spelling Greek words. But as its original place had been filled up by g, it was tacked on as an appendage, rather than incorporated as an element. X in _power_, coincided with the Greek xi; in _place_, with the Greek _khi_. Its _position_ seems to have determined its _form_, which is certainly that of X rather than of Ξ. The full investigation of this is too lengthy for the present work. § 165. It should be observed, that, in the Latin, the letters have no longer any _names_ (like _beth_, _bæta_), except such as are derived from their powers (_be_, _ce_). § 166. The principles which determined the form of the Roman alphabet were, upon the whole, correct; and, hence, the Roman alphabet, although not originally meant to express an Italian tongue at all, expressed the language to which it was applied tolerably. On the other hand, there were both omissions and alterations which have had a detrimental effect upon the orthography of those other numerous tongues to which Latin has supplied the alphabet. Thus--