to alphabetic characters, and here the letters resemble those of the Láts, or of the Caves in the West of India, the most ancient written form of the Sanskrit language.

[In further illustration of the symbols impressed upon this crude money, I insert copies of some novel emblems that occur on a similar class of primitively-adjusted weights of silver in the collection of the Asiatic Society of Bengal. These specimens of the introductory phase of the moneyer's art seem to have been derived from a different section of the continent of India to those that have hitherto been presented to the public; some of their devices probably shadow forth the emblems of a seaboard kingdom; but the varieties are otherwise interesting as showing that the system of this simply-formed currency was as wide-spread, as its products are proved to have been ample.

M

४०४

With a view to test the authoritative issue-weight of these '

1 [This is probably the most fitting place to insert a brief note on ancient Indian weights for gold and silver, condensed from Colebrooke's paper on the subject in the Asiatic Researches.' The lowest division of the Hindú scheme is the trasarenu or ransi defined as a 66 very small mote which may be discerned in a sun-beam passing

punch-marked pieces of silver, I have examined in detail the extended series of specimens in the British Museum, and selected for trial only those examples that appeared to be un-abraded through a lattice: it is further described by medical writers as comprising 30 paramánu or atoms."

3 yava

=

1 liksha, or ‘minute poppy-seed.'

1 rája sher-shapa, or black mustard-seed.'

1 gaura sher-shapa, or white mustard-seed.'
1 yava, or 'middle-sized barley-corn.'

1 krishnala, or 'seed of the ganja.'

This last is the lowest denomination in general use, and commonly known by the name of ratti, (rattiká) which denotes the red seed, as krishnala indicates the black seed, of the ganjá creeper. The jewellers' rati is rated as = ths. of a carat. The average, taken from the seeds themselves, according to Sir W. Jones, gives 1 grain and ths.

A rattiká is also said to be equal to 4 grains of rice in the husk, and by others is stated to correspond with two large barley-corns. Mr. Colebrooke goes on to remark, "Notwithstanding this apparent uncertainty in the comparison of a seed of the ganja to other productions of nature, the weight of a raktika is well determined by practice, and is the common medium of comparison for other weights."

1 dharana or purana.
1 satamána or pala.

A karsha, or 80 raktikás, of copper is called a pana or kársha-pana.

Four varieties of máshas are also noticed, three of 5, 4 and 16 raktikás respectively, and a fourth (the másha of silver) consisting of 2 raktikas. The author proceeds to note further modifications of these weights, which need not be cited in detail; but it is important to mark the definition of the dinár, as the term will be found to be employed in some of the ancient inscriptions referred to in the body of this work. "108 suvernas or tolakas of gold are said to constitute an urubhúshana, pala, or dinára."

Gopala Bhattá also states other weights, without mentioning by what classes they are used. I suspect an error in the statement, because it reduces the másha to a very low denomination, and I suppose it to be the jewellers' weight.

6 rájikás (raktikás)

4 máshas

2 tankas

2 konas

1 másha, hema or vánaka.
1 tanka, sala, or dharana.
1 kona.

1 karsha.

Probably it should be raktikás instead of rájikás, which would nearly correspond with the weights subjoined, giving 24 rettikás for one dharana in both statements. It also corresponds with the Tables in the 'Ayín-i Akberi' (iii., 94,) where a tank of 24 rettis, fixed at 10 barley-corns to the retti, contains 240 barley-corns; and a másha of 8 rettis, at 7 barley-corns each, contains 60 rettis; consequently, 4 mashas are equal to one tanka, as in the preceding table; and 6 jewellers' rettis are equal to 8 double rettis as used by goldsmiths. The same author (Gopala Bhattá) observes that weights are thus stated in astronomical books:

2 large barley-corns 1 seed of the gunja.

3 gunjás

8 ballas

2 dharanas

1000 alakas

and un-oxydized; the result of the weighment of ten such samples gives an average of 52-98 grains, the highest total being 54, the lowest 52 grains,-while four specimens out of the ten were exactly balanced at 53 grains. Of the circular, or rounded-off specimens, the average was somewhat lower, but many of these equally came up to the 53 grains; the metal of which they are composed is, however, clearly debased below the standard of their less advanced prototypes.

They take the shape of ths. byths. of an inch, half-an-inch, and are occaUnlike the cognate series,

Next presents itself a series of sections of metal, which may either be of earlier or of later date than the above. They appear more primitive in their crudity of make, but seem more definitively developed in their stamp. mere divisions of a bar of silver, into lengths of something less than sionally rounded off at the ends. they are marked by a single stamp only, which occupies one superficies and represents two connected balls, forming probably such portion of the die, seen in No. 33, as their surfaces were broad enough to receive the impress of. The best specimen of this somewhat rare class weighs 24.5 grains.

And here I would notice a considerable advance in the art of coining, and one that successfully illustrates the independent local progress, reminding us of the parallel incused-square of many of the early Greek medals. In this instance, the square die is driven home upon one surface so as to give the coin a concavo-convex form, while the opposite face remains blank.'

The tale of shells, compared to weight of silver, is quoted as follows from the 'Lilavati:' = 1 kákíní.

20 kapardakas (shells, cowries)

1 pana, kárshapana, or kúrshika.
1 bherma of silver.
= 1 nishka of silver.

=

It may be inferred that one shell is valued at 1 raktiká of copper; one pana of shells at 1 pana of copper; and 64 panas at one tolaka of silver, which is equal in weight to one pana of copper. And it seems remarkable that the comparative value of silver, copper, and shells, is nearly the same at this time (1799) "as it was in the days of Bhaskara."- Asiatic Researches,' v., 91.]

[Though, in some instances, it would almost seem as if it had been designed to produce a crude device on the convex surface, as is common in many of the Greek examples.]

The symbol within the square has the appearance of a rude quatrefoil. These coins are, as far as I know, unpublished: I have never met with any in the entire course of my own somewhat comprehensive search after local antiquities, nor am I aware from what section of the country the examples I quote were obtained; they now form part of the late Lord Auckland's collection in the British Museum. They are composed of silver considerably alloyed, and weigh from thirty-five up to forty-one grains.

The other silver varieties of early date seem to be sufficiently illustrated by Prinsep's text, so, without further interruption, I resume the sequel of his observations. ']

From the above original seem to have descended two distinct families, of which one was produced by the hammer and die, the other by casting in a mould. Of the latter-easily recognizable by the depth of relief; the projecting keel on the margin, shewing where the moulds were united; and the greater corrosion due to the softness of the cast metal;--we have various groups and sub-divisions, but most of them agree in bearing the Chaitya for their obverse; sometimes-as in figs. 34, 35, 36, 37-with the addition of two smaller symbols, like the sign of Taurus reversed.

On the reverse, we have frequently a dog with a collar (and bell?) guarding a sword or 'flag-staff of victory,' (jaya-dhwaja?) figs. 20, 21, 34, 35, 36. At other times an elephant (fig. 39); a bull (37), or the sacred-tree (15, 38); and, in rarer cases, the device on

1 [I take this opportunity of referring to certain punch-marked gold coins in Prinsep's Collection in the British Museum. These seem to be of a far more modern date than the specimens of the silver series: they are impressed with punch dies, bearing devices of a more recent character and execution, and usually display a superstruck legend consisting of four characters only partially intelligible. The locality of their discovery is not noted; the weights of the four specimens are as follow: 57.5, 580, 58.3, and 58.3.]

both sides is changed, as in 40, 41. Figs. 18, 42, and 43, (in the latter of which the elephant might easily be mistaken for a Deva-nágarí letter,) are of the cast species; to which also belongs the multi-symbolic coin, fig. 18 of the last plate, and its fellows of former plates. The leaden coin, 49, is also cast, but it is probably a forgery from some copper original.

[Further examples and varieties of the general series of cast coins have since come to light, to which I refer here merely to exemplify how diversified both the types and fabric of these early pieces were, and the time that, in consequence, must reasonably be conceded to have elapsed between the commencement and the termination of this system of coinage. We have no sufficient data for separating the locality of their issue from the ordinary field of the collateral mintage of die-struck coins; for, as far as can be inferred from the sites of discovery, they would almost appear to have been parallel improvements upon the primitive punch-marked weights of metal: that they progressed in their own independent development, may however be admitted, from the advanced character of the outward form, and the superior treatment of the devices, of the later specimens.

As might be expected, they altogether fail to afford any proximate solution of the intentional standard under which they were put forth, as the weights vary to a marked degree. Wellpreserved examples of No. 37, now in the British Museum, range from 49 to 64, 65, and 68 grains, while specimens of No. 34 vary from 64 to 74 grains.

I do not propose at present to enter into any minutia in illustration of the various classes of cast coins, but I would refer to a somewhat curious division of the manufacture, in the shape of coins cast in copper or bronze, with the one side left altogether blank. I infer that Nos. 30, 31, pl. xv., Ariana Antiqua,' are of this description, though Prof. Wilson does not notice the method of formation.