(3.) We now turn to collateral sources of information, which we will follow out, as far as possible, in chrono- logical order. The earliest and most trustworthy testi- mony as to the length of the cubit is supplied by the existing specimens of old Egyptian measures. Several of these have been discovered in tombs, carrj-ing us back at all events to B.C. 1700, while the Kilometer at Elephantine exhibits the length of the cubit in the time of the Roman emperors. No great difference is exhib- ited in these measures, the longest being estimated at about 21 inches, and the shortest at about 20i, or ex- actly 20.4729 inches (Wilkinson, .4 nc. Kg. ii, 258). They are divided into 28 digits, and in this respect contrast with the Mosaic cubit, which, according to rabbinical authorities, was divided into 24 digits. There is some difiiculty in reconciling this discrepancy with the almost certain fact of the derivation of the cubit from Egypt. It has generally been surmised that the Egyptian cubit was of more than one length, and that the sepulchral measures exhibit the shorter as \\c\\ as the longer by special marks. Wilkinson denies the existence of more tiian one cubit {Anc. Kg. ii, 257-259), apparently on the gniuiid that the total lengths of the measures do not materially vary. It may be conceded that the measures are intended to represent the same length, the variation being simply the result of mechanical inaceiu-acy ; but this does not decide the question of the double cubit, which rather turns on the peculiarities of notation ob- servable on these measures. For a full discussion of this point we must refer the reader to Thenius's essay in the Theologische Sludien mid Kriliken for 184G, p. 297-342. Our limits will permit only a brief statement of the facts of the case, and of the views expressed in reference to them. The most perfect of the Egyptian cubit meas- ures are. those preserved in the Turin and Louvre mu- seums. These are imequally divitled into two parts, the one on the right hand containing 15, and the other 13 digits. In the former part the digits are subdiWded into alitpiot parts from ^ to J,, reckoning from right to left. In the latter part the digits are marked on the lower edge in the Turin, and on the upper edge in the Louvre measure. In the Turin measure the three left-hand digits exceed tlie others in size, and have marks over them indicating either fingers or the numerals 1, 2. 3. The four left-hand digits are also marked off from the rest by a double stroke, and are further distinguished by hieroglyi)hic marks sujiposed to indicate thai tliey are digits of the old measure. There are also special marks between the 6th and 7th, and between tlie 10th and 11th digits of the left-hand portion. In the Louvre cubit two digits are marked off on the lower edge by lines running in a slightly transverse direction, thus producing a greater length than is given on the upper side. It has been found that each of the three above specified digits in the Turin measure = ^j'-j of the whole length, less these three digits; or, to jnit it in another form, the four left-hand digits = Jj of tlic 25 right-hand digits: also that each of the two digits in the Louvre measure — ;_?;j of the whole length, less these two digits; and further, that twice the left half of cither measure =^
METROLOGY
195
METROLOGY
the whole length of the Louvre measure, less the two digits. Most writers on the subject agree in the con- clusion that the measures contain a combination of two, if not three, kinds of cubit. Great difference of opinion, however, is manifested as to particulars. Thenius makes the difference between the royal and old cubits to be no more than two digits, the average length of the latter being 484.289 millimetres, or 19.0(5G inches, as compared with 523.524 millimetres, or 20.611 inches, and 523 mil- limetres, or 20.591 inches, the lengths of the Turin and Louvre measures respectively. He accounts for the ad- ditional two digits as originating in the practice of placing the two fingers crosswise at the end of the arm and hand used in measuring, so as to mark the spot up to which the cloth or other article has been measured. He further finds, in the notation of the Turin measure, indications of a third or ordinary cubit 23 digits in length. Another explanation is that the old cubit con- sisted of 24 or 25 new digits, and that its length was 462 millimetres, or 18.189 inches; and, again, others put the old cubit at 24 new digits, as marked on the meas- ures. The relative proportions of the two would be, on these two hypotheses, as 28 : 26, as 28 : 25, and as 28 : 24. (See below.)
The use of more than one cubit appears to have also prevailed in Babylon, for Herodotus states that the " royal" exceeded the '• moderate" cubit (w/jx^c /'* Tp'of) by three digits (i, 178). The appellation "royal," if borrowed from the Babylonians, would itself imply the existence of another; but it is by no means certain that this other was the •' moderate" cubit mentioned in the text. The majority of critics think that Herodotus is there speaking of the ordinary Greek cubit (Biickh, p. 214), though the opposite view is affirmed by Grote in liis notice of Biickh's work {Claxs. Mits. i, 28). Even if the Greek cubit be understood, a further difficulty arises out of the uncertainty whether Herodotus is speaking of digits as they stood on the Greek or on the Babylonian measure. In the one case the proportions of the two would be as 8 : 7, in the other case as 9 : 8. Bockh adopts the Babylonian digits (without good rea- son, we think), and estimates the Babylonian roval cubit at 234.2743 Paris lines, or 20.806 inches (p. 219). A greater length would be assigned to it according to the data furnished by M. Oppert, as stated in Rawlinson's Ilefod. i, 315 ; for if the cubit and foot stood in the ratio of 5 : 3, and if the latter contained 15 digits, and had a length of 315 millimetres, tlien the length of the ordi- nary cubit would be 525 millimetres, and of the royal cubit, assuming, with Mr. Grote, that the cubits in each case were Babylonian, 588 millimetres, or 23.149 inches.