"Madam,—I writ this Letter two months ago, and was to send it by Mrs. Barber; but she falling ill of the gout, and I deferring from day to day, expecting her to mend, I was at last out of patience. I have sent it among others by a private hand.

"I wish Your Ladyship and all your family many happy new years.

"Jan. 8e, 1732."

NINEVEH INSCRIPTIONS.

The accumulation of these treasures in London and Paris, leads to the belief that they will soon be decyphered. The following remarks are offered in promotion of so desirable an object. It must be premised that a printer, when requiring type from the type-founder for English books, does not order the same quantity for each letter; but, according to a scale adapted to the requirements of printing, he orders only so many of each letter as he is likely to use. That scale may be nearly represented in the following way: the letter z being the one least used in English, he will require

Twice the number of letter z for letter x
Twice also —— j
2½ times —— q
4 " —— k
6 " —— v
8 " —— b
8½ " —— p
8½ " —— g
10 " —— y
10 " —— w
15 " —— m
15 " —— c
17 " —— u
20½ " —— l
21 " —— f
22 " —— d
31 " —— r
32 " —— h
40 " —— s
40 " —— n
40 " —— o
41½ " —— i
42½ " —— a
45 " —— t
60 " —— e

Suppose now a person to write English in cypher, using unknown characters for the well-known letters; it would be easy to decypher his writing, if of sufficient length to make the general rule acted on in the printing trade applicable. The decypherer, by selecting each distinct unknown character, and numbering them respectively, would find that the character oftenest occurring was e, the next oftenest t, and so on to the character having the lowest number, being least used, which would of course be z. Persons accustomed to decypher European correspondence for diplomatic purposes, will pronounce best on the practicability of this method for the decyphering of modern languages.

It is proposed then to apply the same method in the several languages supposed nearest of kin to that of the Nineveh inscriptions. Without entering into the reasons for that opinion, it may suffice, for the present purpose of illustration, to assume that the language of these inscriptions is Chaldee. To apply this method the numbers of each letter occurring in the Targum of Onkelos on Genesis, or the whole Pentateuch, should be taken. This enumeration has been made as regards the Hebrew (see Bagster's Family Bible, at the end of Deuteronomy). The readiest mode of effecting such enumeration would be to employ twenty-two persons knowing the Chaldee letters, and to assign a letter to each, calling out to them each letter as it occurred in Onkelos, whilst each person kept count of his own letter on a tally, and summing up the total gave in the result to the reader at the end of each chapter. This would be necessary with a view to ascertain what quantity of unknown inscription was required to evolve the rule, as the proposed method is clearly inapplicable when the quantity of matter to be decyphered is inconsiderable.

Having gone over sufficient ground to satisfy himself of the certainty of the rule, the decypherer would next count the numbers of each distinct character in all the cuneiform inscriptions accessible to him, making allowance for final letters, also for vowel points which may be attached to the character, as in Ethiopic. Assuming the rule in Chaldee to be the same as in Hebrew (it is in fact very different), he would find the character oftenest occurring in the Nineveh inscriptions to be ו, the next מ, the rest in the following order as to frequency of occurrence, ט , ס , ע , צ , [?] ד , פ , ז , ק , [?] ח , [?] ב , ש , [?] ד , [?] ב , ל , נ , א , ה , כ , ת , י , the first letter, ו, vau, occurring nearly seven times as often as ט, teth. The order of the letters would, in fact, vary much from this in Chaldee; the servile letters being different would alone much disturb the assumed order, actually ascertained nevertheless, as respects the Hebrew letters, in the five books of Moses. One word as to the order in which the several languages should be experimented on. The Chaldee would be the first, and next in succession, (2) the Syriac, (3) the Ethiopic, (4) the Arabic, (5) the Hebrew (die jungste Schwester),[3] and (6) the Pehlvi. The Indo-European languages would, in case of failure in the above, claim next attention: of these first the Zend, next (2) the Sanscrit, then (3) the Armenian, &c. &c.