I.d and the Greek Δ, was the fourth letter of the Latin alphabet, and was called de: Ter. Maur. p. 2385 P., Auson. Idyll. 12, de Litt. Monos. 14. But at the end of a syllable, or after another consonant, its sound was sharpened, so that the grammarians often discuss the question whether d or t should be written, especially in conjunctions and prepositions. Illa quoque servata est a multis differentia, ut ad cum esset praepositio, d litteram, cum autem conjunctio, t acciperet (Quint. 1, 7, 5; cf. id. 1, 4, 16). Hence we may infer that some disputed this distinction, and that the sounds of ad and at must at least have been very similar (cf. also Terent. Scaur. p. 2250, Vel. Long. p. 2230 sq., Cassiod. p. 2287, 2291). Thus also aput, it, quit, quot, aliut, set, haut are found for apud, id, quid, quod, aliud, sed, haud. It would appear from the remarks of these authors that the last two words in particular, having a proclitic character, while they distinctly retained the d sound before an initial vowel in the following word, were pronounced before a consonant almost as set, haut (Mar. Vict. p. 2462 P., Vel. Long. l. l. v. Corss. Ausspr. 1, 191 sq.). The use of t for d in the middle of a word, as Alexenter for Alexander, atnato for adnato, is very rare (cf. Wordsworth, Fragm. p. 486 sq.). On the other hand, the use of d for t, which sometimes appears in MSS. and inscrr., as ed, capud, essed, inquid (all of which occur in the Cod. palimps. of Cic. Rep.), adque, quodannis, sicud, etc., fecid, reliquid, etc. (all in inscriptions after the Augustan period), is to be ascribed to a later phonetic softening (cf. Corss. Ausspr. 1, 191 sq.).
II. As an initial, the letter d, in pure Latin words, suffers only a vowel after it; the single consonantal compound dr being found only in borrowed words, such as drama, Drusus, Druidae, etc., and in the two onomatopees drenso and drindio. Accordingly, the d of the initial dv, from du, was rejected, and the remaining v either retained unaltered (as in viginti for duiginti; cf. triginta) or changed into b (as in bellum, bis, bonus, for duellum, duis, duonus; v. those words and the letter B). So too in and after the 4th century A.D., di before vowels was pronounced like j (cf. Jovis for Djovis, and Janus for Dianus); and hence, as the Greek δι (di) passed into dz, i. e. ζ (as in ζ α for δ ια, and zeta for diaeta), we sometimes find the same name written in two or three ways, as Diabolenus, Jabolenus, Zabolenus; Jadera, Diadora, Zara. In many Greek words, however, which originally began with a y sound, d was prefixed by an instinctive effort to avoid a disagreeable utterance, just as in English the initial j has regularly assumed the sound of dj: thus Gr. ζυγόν, i. e. διυγον = L. jugum; and in such cases the d sound has been prefixed in Greek, not lost in Latin and other languages (v. Georg Curtius Griech. Etym. p. 608 sq.).β. As a medial, d before most consonants undergoes assimilation; v. ad, no. II.; assum, init., and cf. iccirco, quippiam, quicquam, for idcirco, quidpiam, quidquam; and in contractions like cette from cedite, pelluviae from pediluviae, sella from sedela. In contractions, however, the d is sometimes dropped and a compensation effected by lengthening the preceding vowel, as scāla for scand-la. D before endings which begin with s was suppressed, as pes from ped-s, lapis from lapid-s, frons from frond-s, rasi from radsi, risi from rid-si, lusi from lud-si, clausi from claud-si; but in the second and third roots of cedo, and in the third roots of some other verbs, d is assimilated, as cessi, cessum, fossum, etc. D is also omitted before s in composition when another consonant follows the s, as ascendo, aspicio, asto, astringo, and so also before the nasal gn in agnatus, agnitus, and agnosco, from gnatus, etc.: but in other combinations it is assimilated, as assentio, acclamo, accresco; affligo, affrico; agglomero, aggrego; applico, approbo, etc. In tentum, from tendo, d is dropped to avoid the combination ndt or ntt, since euphony forbids a consonant to be doubled after another.γ. Final d stood only in ad, apud, sed, and in the neuter pronouns quid, quod, illud, istud, and aliud, anciently alid. Otherwise, the ending d was considered barbarous, Prisc. p. 686 P.
III. The letter d represents regularly an original Indo-Germanic d, in Greek δ, but which in German becomes z , in Gothic t, and in Anglo-Saxon t: cf. Gr. ἥδομαι, Sanscr. svad, Germ. süss, Angl.-Sax. svēte (sweet), with Lat. suadeo; domare with Gr. δαμάω, Germ. zähmen, Eng. tame; domus with δέμω, timber, O. H. Germ. zimber; duo with δύω, zwei, two. But it is also interchanged with other sounds, and thus sometimes represents—
1. An original t: mendax from mentior; quadraginta, quadra, etc., from quatuor.—
2. An original r: ar and ad; apur or apor and apud; meridies and medidies, audio and auris; cf. arbiter, from ad-beto; arcesso for ad-cesso.—
3. An original l: adeps, Gr. ἄλειφα; dacrima and lacrima, dingua and lingua; cf. on the contrary, olere for odere, consilium and considere, Ulixes from Ὀδυσσεύς (v. Corss. Ausspr. 1, 223).—
4. An original s: Claudius, from the Sabine Clausus, medius and μίσος; and, on the contrary, rosa and ῥόδον. —
IV. In the oldest period of the language d was the ending of the ablat. sing. and of the adverbs which were originally ablatives (cf. Ritschl, Neue Plaut. Excur. I.; Brix ad Plaut. Trin. Prol. 10): puCNANDO, MARID, DICTATORED, IN ALTOD MARID, NAVALED PRAEDAD on the Col. Rostr.; DE SENATVOS SENTENTIAD (thrice) IN OQVOLTOD, IN POPLICOD, IN PREIVATOD, IN COVENTIONID, and the adverbs SVPRAD SCRIPTVM EST (thrice), EXSTRAD QVAM SEI, and even EXSTRAD VRBEM, in S. C. de Bacch. So intra-d, ultra-d, citra-d, contra-d, infra-d, supra-d; contro-d, intro-d, etc.; and probably interea-d, postea-d. Here too belongs, no doubt, the adverb FACILVMED, found in the last-mentioned inscription. But this use of the d became antiquated during the 3d century B.C., and is not found at all in any inscription after 186 B. C. Plautus seems to have used or omitted it at will (Ritschl, Neue Plaut. Excurs. p. 18: Corss. Ausspr. 1, 197; 2, 1008).
2. D final was also anciently found—
a. In the accus. sing. of the personal pronouns med, ted, sed: INTER SED CONIOVRASE and INTER SED DEDISE, for inter se conjuravisse and inter se dedisse, in the S. C. de Bacch. This usage was retained, at least as a license of verse, when the next word began with a vowel, even in the time of Plautus. But in the classic period this d no longer appears. —
b. In the imperative mood; “as estod,” Fest. p. 230. The Oscan language retained this ending (v. Corss. Ausspr. 1, 206).—
c. In the preposition se-, originally identical with the conjunction sed (it is retained in the compound seditio); also in red-, prod-, antid-, postid-, etc. (redire, prodire, etc.); and in these words, too, it is a remnant of the ancient characteristic of the ablative (v. Corss. Ausspr. 1, 200 sq.; Roby, Lat. Gr. 1, 49).
V. As an abbreviation, D usually stands for the praenomen Decimus; also for Deus, Divus, Dominus, Decurio, etc.; over epitaphs, D. M. = Diis Manibus; over temple inscriptions, D. O. M. = Deo Optimo Maxumo; in the titles of the later emperors, D. N. = Dominus Noster, and DD. NN. = Domini Nostri. Before dates of letters, D signified dabam, and also dies; hence, a. d. = ante diem; in offerings to the gods, D. D. = dono or donum dedit; D. D. D. = dat, dicat, dedicat, etc. Cf. Orell. Inscr. II. p. 457 sq.!*? The Romans denoted the number 500 by D; but the character was then regarded, not as a letter, but as half of the original Tuscan numeral (or CIↃ) for 1000.