the manner in which in the early times the year was divided and defined, we shall commence at once with the Julian year and its division into months. According to this the month of February in a common year had twenty-eight days; April, June, September, and November thirty, and the others thirtyone days. The days of these months are not reckoned in an uninterrupted series, from one to thirty or thirty-one, but are calculated backwards from three days, which are fixed in every month. These three days are the first, fifth, and thirteenth, which are called by their Roman names the Calendae, Nonae, and Idus, of a month. (The names of the months, as was remarked in § 38., are used as adjectives, and as such they are joined to the three feminine names just mentioned.) In the Roman system of counting from a certain point, this point itself is included in the calculation. Thus, e. g. the third day before the nonae, i. e. before the fifth of the month, is not the second of the month, but the third. Hence we may give it as a practical rule, that in calculating the days of the month, we must add one to the number from which we deduct. When the point from which we have to count backward is the first of the month (Calendae), it is not sufficient to add one to the number of days of the current month, but the Calendae itself must also be taken into the account, i. e. the number of days of the current month must be increased by two before we deduct from them. Hence dies tertius ante Cal. Julias is the 29th of June, as June has thirty days. This is the cause of the whole apparent difficulty in calculating the Roman dates. But besides this we have to consider another peculiarity, which is a remnant of the ancient arrangement of the Roman year, ascribed to king Numa, viz. in the months of March, May, July, and October, the Nonae fall on the 7th, and the Idus on the 15th, instead of the 5th and 13th. In leap-years (i. e. according to the Roman expression every fifth year) February has one day more, but this intercalary day was not added at the end of the month, as is the custom in modern times, but was inserted in the place where formerly the intercalary month (mensis intercalaris) had been inserted to make the lunar year of king Numa harmonise with the solar year, that is, after the 23d of February, so that the 24th of February, i. e. the sixth day before the Calendae of

March, was reckoned double, and was called bis sextus or bis sextum, whence the leap-year itself was called annus bis sextus. On this subject see the classical work of Ideler, Handbuch der mathematischen und technischen Chronologie, Berlin, 1825, in the beginning of vol. ii.

be

[§ 868.] Respecting the grammatical form of stating the day of a month the following points must be observed. The ablative indicates the time when a thing occurs; hence we say, e. g. die tertio ante Calendas Martias, but die and ante may omitted, and we may say tertio Calendas, or in figures iii. Cal. Cicero and Livy, however, use a different form either exclusively, or at least much more commonly than others; e. g. ante diem tertium Calendas, or Nonas, Idus (abridged a. d. iii. Cal.). This peculiarity, instead of the correct die tertio ante Calendas, cannot be explained otherwise than by the supposition that ante changed its place, and that afterwards the ablat. was changed into the accusat., as if it were dependent on ante; while the real accusat. Calendas remained unchanged. Pridie, the day before, and postridie, the day after, are either joined with the genitive, as pridie ejus diei, or in the case of established calendar names and festivals, with the accusative, to which people were more accustomed, as pridie Idus, pridie Compitalia, pridie natalem, postridie ludos Apollinares.

[§ 869.] This expression ante diem must be considered as an indeclinable substantive, since we often find it preceded by prepositions which govern the accusat. or ablat., e. g. Cic. in Cat. i. 3. dixi ego idem in Senatu, caedem te optimatum contulisse in ante diem V. Cal. Novembris (or Novembres, is being probably only the ancient termination of the accusat., instead of es); Liv. xliii. 16.: in ante dies octavum et septimum Calendas Octobres comitiis dicta dies; xlv. 2.: supplicatio indicta est ex ante diem quintum Idus Octobres, cum eo die in quinque dies; and in the same manner postridie, e. g. Cic. ad Att. ii. 11.: nos in Formiano esse volumus usque ad pridie Nonas Majas.

[§ 870.] In order to facilitate the calculation of a date in the ancient calendar (such as it was established by C. Julius Caesar, in B. C. 45), we have annexed Bröder's table, in which the beginner may easily find his way.

ROMAN WEIGHTS, COINS AND MEASURES.

[§ 871.] 1. THE Roman pound (libra, pondo) is about of the Paris pound, that is 11 ounces and 1 drachm. (According to Romé de l'Isle it contained 6048 Paris grains, according to Ca

gnazzi 6135, according to Letronne 6154, according to Paucker and Böckh 6165, 9216 of which make a Paris pound.) It is divided into 12 parts (unciae), and these twelve parts together are called an as. The names of the fractions are: is uncia (about an ounce in weight);

half an as;

three or;

septunx;

sextans, that is,

quincunx;

8

; quadrans, that is, semis or semissis ; i. e.

bes or bessis, i. e. two parts out of dodrans, compounded from dequadrans, i. e. §; 12 dextans or decunx; ¦¦ deunx, i. e. one ounce less, scil. than an as. These names are also applied to other relations; thus we say; e.g. he was instituted heir ex dodrante; i.e. he received ; ex deunce, he received of the whole property. An uncia contains 2 semiunciae, 3 duellae, 4 sicilici, 6 sextulae, 24 scrupula or scripula. One ounce and a half is sescuncia (from sesquiuncia). Compounds of as are tressis, 3 ases; octussis, 8 ases; decussis, 10 ases; centussis, 100 ases.

9

[§ 872.] 2. The most ancient Roman money was of copper, and the as as a coin, was originally a pound of copper coined. At the time when the Romans commenced to coin silver (some years before the first Punic war), the copper as was reduced, at first to, afterwards to, and at last to of the original weight, so that the coin which had originally weighed a pound of copper, was afterwards only half an ounce in weight.

Silver coins were the denarius, originally equal to 10 ases, and subsequently, after the reduction of the as to, equal to 16 ases. Half a denarius was called quinarius; of a denarius sestertius, that is, originally 2 ases and a half (hence it is written HS; i. e. 2), but when the denarius had become equal to 16 ases, it was worth 4 ases. Silver coins of still smaller value were the libella, = of a denarius; the sembella, = ' of a denarius; teruncius, of a denarius, 3 unciae of the ancient, and 4 unciae of the reduced copper money. A denarius weighed a little more or less than 73 Paris grains, but was gradually reduced, under the first emperors, to 63 grains; hence the Roman pound in the times of the republic contained about 84 denarii (which, according to Plin. Hist. Nat. xxxiii. 46., was the legal amount), and in the reign of Domitian from 96 to 100.

Gold was coined in various ways: an aureus in the times of the emperors was equal to 25 denarii or 100 sestertii; conse

quently 1000 HS are equal to 10 aurei, 100,000 HS to 1000 aure, and decies HS to 10,000 aurei. The emperor Honorius made 25 pounds of copper coin equal to one solidus (aureus), that is, a pound of copper equal to a silver denarius.

[§ 873.] 3. The Romans generally calculated according to sestertii, and a nummus is simply a sestertius. Instead of mille sestertii, we may say with equal correctness, mille sestertium (genit. plur.), just as we commonly say mille passuum. A million, as was remarked in § 115., is expressed by the form of multiplication: decies centena milia sestertium, or more commonly by decies alone, centena milia being omitted; centies therefore is 10 millions, and millies 100 millions. As people were thus accustomed to hear the word sestertium in connection with mille, they came by a kind of grammatical blunder to consider sestertium as a substantive of the neuter gender, and hence they said unum sestertium, septem sestertia, bis dena sestertia, sexcenta sestertia, &c., instead of unum mille sestertium, septem milia sestertium, &c. In Cicero it does not often occur, but is yet found in some passages, as in Verr. iii. 50.; Parad. 6. 3.; but in the writers of the silver age it is quite common.

Decies sestertium, a million of sestertii, centies sestertium, &c. is used as a singulare tantum of the neuter gender; e.g. Cic. in Verr. ii. 7.: HS decies numeratum esse; Philip. ii. 16. amplius HS ducenties acceptum hereditatibus rettuli. But the mistake was carried still further by declining this expression; e. g. Liv. xlv. 4. argenti ad summam sestertii decies in aerarium rettulit, up to the sum of one million sestertii; Cic. Philip. ii. 37.: syngrapha sestertii centies, a bill of ten millions of sestertii; Tacit. Ann. xii. 58. Bononiensi coloniae, igni haustae, subventum centies sestertii largitione, by a present of ten millions of sestertii; Sueton. Caes. 50.: Serviliae sexagies sestertio margaritam mercatus est, he bought her a pearl for six millions of sestertii; Sueton. Octav. 41.: Senatorum censum duodecies sestertio taxavit, he fixed the senatorial census at 1,200,000 sestertii; Cic. p. Font. (Niebuhr, Fragm.) § 4.: Testis non invenitur in ducentis et tricies sestertio; ad Att. iv. 2.: superficiem aedium aestimarunt HS (sestertio) decies.

[§ 874.] 4. With regard to Greek weights and money, we can here add only a few remarks. An Attic talent (talentum) is equal to 80 Roman pounds; a mina (uvâ) is the sixtieth part