ever-growing language too; and those who spoke it could often express in a single word ideas which it costs us a whole string of words to convey. Consequently the growth of Latin words from their primary and secondary roots would be most clearly exhibited by means of a genealogical tree such as the foregoing; which, it will be observed, includes many of the derivatives of the supposed root pu- that do not occur in Virgil, and which are not found in the Vocabulary, but are here added in order to set forth the process with some degree of completeness. * 9. It is plain that such a plan as this, if it could be thoroughly carried out, would give at a glance full and most valuable information as to the history of Latin words, and the exact degree of their relationship to their root and to each other, whether as direct descendants (like children and grandchildren), or as derived from one common ancestor (as cousins are); and this full knowledge of a word's history would be of the greatest use in determining its exact force and meaning. The chief drawback to this plan is the room it would take, especially if the declensions, &c., and English meanings were added to the Latin words. 10. I have attempted therefore in the Vocabulary to convey the same information in more convenient form by the careful use of thick type. This thick type is employed in the first instance to distinguish the supposed root-letters of the first word of the group; and then in the later words to mark the part that they have borrowed from the words of older standing than themselves, and serving as roots to them, or secondary roots, as they are commonly called, in order to distinguish them from primary or original roots. 11. Thus the primitive root runs like a back-bone through the whole of each group, giving it firmness and connection; whilst the precise meaning of each individual word in the group is determined by its own suffixes and those of its immediate ancestors. A short study of the genealogical tree above given will show how the dark letters of the root-type gradually increase in number at each step of direct descent, * See preceding page. whilst they remain the same in words which stand on the same line or level, the brothers and sisters, as it were, in each particular family. 12. It is necessary however to observe that root-letters often underwent certain changes for the sake of harmony of sound; and this law of harmony, or gratification of the ear, seems to have worked in two opposite ways. For first there is the well-known law of assimilation, according to which a consonant is often changed into the same letter that immediately follows it, so that ad+ferre becomes afferre; or it at least takes the same character as in Greek Ev+кóπтw becomes ÉуKÓTT. But on the other hand, if the two consonants are separated by a vowel which cannot be dropped, a law of repulsion takes the place of the law of attraction, and a change takes place for the sake of variety, and thus medio+dies, or medidies, becomes meridies, and coeruleus probably takes the place of coeluleus; and so in Greek we have nom. Opíž, gen. Tpxós; and in pl. gen. Tpixwv, dat. Opiči, the one aspirate coming in where the other disappears, as is also the case with τύφθητι and τύπηθι; τρέφω, θρέψω, &c. And in the choice of suffixes the same rule prevails, so that puer becomes puerilis, whilst puella makes its adj. puellaris, the -ris and -lis being identical in meaning, and the letter being chosen which does not already exist in the word. The subject cannot be further discussed here; but in the Vocabulary the representatives of root-letters have received the same dark type as the rootletters themselves would have done if they had remained unaltered. 13. Whilst the dark type has thus been employed to distinguish the root-letters (or, as they are sometimes called, the radicals of Latin words), capitals are used to mark those English words which are connected with their corresponding Latin words in origin as well as in meaning. But this connection of origin is of two perfectly distinct kinds; and the distinction is one which deserves careful attention. For in the first place very many English words are actually derived from the Latin either directly, as our word SCRIPTURE is drawn straight from the Latin scriptura (a writing); or else through the French, as our PATERNAL has come to us ultimately indeed from the Latin paternus, but still only at second hand through the French paternel. Accordingly, whether English words are borrowed directly from the Latin (in which case they usually keep their Latin look and spelling) or indirectly (when they are often so disguised on the road as not to be easily recognised at first sight), they are in either case printed in capitals in order to call attention to their Latin origin.* 14. But besides these English words derived from the Latin, there is another large class of English words which ought to be still more interesting to the English student; I mean those words of pure Northern descent, which are clearly related to the corresponding Latin words, and yet not derived from them; and the capital type is used to call attention to these also. For example, FATHERLY is printed in the same large type as PATERNAL; but that is not because it is derived from paternus, nor yet because the English word FATHER is borrowed from the Latin pater (any more than the Latin pater is derived from the English FATHER), but it is because they are both of them alike representatives of one and the same ancient root which has come down to us in these two different forms, the Latin pater acting as the representative of the old languages of Southern Europe; whilst our English FATHER represents those of Northern Europe. The same remark too applies with equal force to such pairs of words as a stable and a stall, &c., where we have borrowed 'stable' from the Latin stabulum; whilst stall, which is our own word for expressing the same idea, comes to us by ancient inheritance from our Gothic, Teuton, German, Saxon, or Old English forefathers; it being a matter of indifference by which of these names we choose to speak of those Northern ancestors, who gave to us * Anyone who wishes to understand the exact process by which Latin words became French (and so in many cases passed on into English) can read a good and plain account of it in Brachet's' Grammaire historique de la Langue française'; where he will also find interesting testimony to the fact that in French there are preserved the remains of an immense number of true Latin words which are not to be found in any Latin writer, because they belonged to the spoken or 'vulgar' tongue and not to the literary dialect. In the same spirit, I once heard a railway guard at Bangor station rebuke a passenger who asked his friend to come into a particular 'box' of the long open carriage-' Compartment you should say, not box; that's so vulgar.' the groundwork and substance of our common speech to this day. 15. The most simple way of accounting for the difference between the northern and the southern forms of kindred words is the strong probability that nations differ as much in the structure of their throats and other organs of speech as they do in the character and expression of their faces; and though the anatomist can give no definite rules to explain the distinction in either case, yet the ear judges of the one difference with just as much certainty as the eye does of the other. 16. At all events it is certain that we have in English many words clearly akin to words of the same meaning in Latin, and just as certainly not borrowed from the Latin; and the special interest of these words arises from the fact that they bear witness to the time, and to the state of society prevailing at the time, when the language of our remote forefathers had not yet branched off into what we may now call its Southern and Northern forms; or into Latin and Greek on the one hand, and the Gothic, German or Old English families on the other. 17. And again, as it is certain that in later times the northern nations of Europe have borrowed many words from its southern races—and even German contains many words of undoubtedly Latin origin-so too it is (to say the least) exceedingly probable that many Latin words were in earlier times taken from northern sources, or at all events the northern tongues retained words which the southern languages lost in their simple form. Thus it comes to pass that we have to this day in our own tongue many words which have come down to us in virtue of our English descent, and which, though not existing in Latin and having no Latin look about them, have nevertheless formed the basis on which Latin words were built up; and then we once more took back our own words from the Latins with the additions which they made to them. For example, this is the easiest way of explaining the connection between such words as hand and apprehend, apprehension, &c. (See 84. B. p. 103.) 18. The capital letters therefore will do good service to the pupil if they lead him to see, first, how much our modern |