As for the Sinhala script (Sinhalese), in Sri Lanka, no national (local) numerals have been known to it throughout two centuries ago - at least; the thing that outweighs they had been using the Arabic numerals in a form of them - like the people of Maldives, then they adopted the ancient Arabic numerals (Ghebari) in transcribing from Portuguese after 1500 AD, as the same has happened by adopting the Portuguese name: "Colombo" for their capital!
The inclination of some Indian nations - on top of them come the Sinhalese - to use the Arabic-Western numerals in their contemporary form in printing, cannot be seen but another proof to the foreignness of those numerals (their local ones) from their scripts; just like nations such as Turkey that has abandoned the Arabic alphabet (which is not national or native to it) in influencing by the western civilization.
Even if we hold a comparison as before between what is alleged to be the first Brahmi numerals, engraved on stones (in the first century AD), and the ancient Brahmi script's alphabet, from which most of the Indian modern scripts have developed, those numerals' belonging will be closer to the Arabic alphabet's letters, yet to the Ghebari numerals (from 5 to 9). That is because such engraved numerals are of a simpler artistic structure and about to be free from points which are frequent in the Brahmi alphabet. Here they are for scrutinizing:

Then you scrutinize the Nabataean alphabet which had been written between 500 BC - 106 AD, as provided by an archeological French source, and represent the origin of the modern Arabic alphabet.

Here they are again after italicizing most of the letters 10 - 45 degrees clockwise, for making the opposition with the modern Arabic letters easier.
Characters' structure is not only what decides the Indian transcribing numerals from Arabs, but the writing-direction too. So by knowing that all European languages read numbers from left to right as, of course, the direction of their reading and writing their scripts, and, for example, the number "35" is read in French, as in almost all European languages: trent cinq, the "thirty" first then the "five", but the German language is an exception to them, since they read the units first, then the tens: fünf und dreiβig (the "five" then the "thirty"). The German orientalist: Dr. Sigrid Hunke, in her book (Allahs Sonne über dem Abendland Unser Arabisches Erbe), has showed that the reason for that is the German being influenced by the Arabs when learnt writing numbers in Arabic numerals.
At the same time, if we know that the non-Moslem Indians talking the Aryan (Indo-European) following languages: Hindi (written in Devanagari), Gujarati (written in Gujarati); Punjabi (written in Gurmukhi) and represent at least 60% of the Indian population - according to the 90's statistics - read the two-digit numbers (21 to 99) from right to left, i.e.: starting with the units then the tens, though all of them write their languages from left to right - for the origination existed between them and the European languages, so under what explanation can the reading of the major Indian nations/languages for the two-digit numbers from right to left be, unless the proprietors and inventors of the numerals are the Arabs? The Indians have learnt them from the Arabs getting influenced by the right-to-left Arabic order of reading numbers, as happened with the German thereafter.
Whereas Arabs not only write and read their language from right to left, but still they, of course, read the numbers from right to left as well and whatever the value to which the numbers amount. As an example, the number 1973 reads correctly in classic Arabic as: three, seventy, nine hundreds, and one thousand. It is worth mentioning that the numbers' reading of the Yemenese: Alsardafi, in his manuscript: "Mokhtasar Alhendi Fi Alhesab مختصر الهندي في الحساب" [The Brief of Alhendi on Arethmatic], such as: "two thousands and five hundreds ألفان وخمسماية" is dissident to the Arabic rule of reading numbers.
On the other hand, the numbers' positions increase from right to left, not from left to right; i.e. decimal positions: units, tens, hundreds, etc. start at the right and continue leftwards, not from left and continue rightwards, while it was fairly possible for their this direction to be otherwise, from left to right, and carrying the arithmetic operations on this basis with no problem at all, so we write and add the "ten" (10), "twenty-seven" (27), and "two hundreds and eighty-three" (283), from the left as follows:
| ٠١ | 01 | |
| + | ٧٢ | 72 |
| + | ٣٨٢ | 382 |
| ||
| = | ٠٢٣ | 023 |
So the sum of those three numbers be: "three hundreds and twenty". The logic necessitates that the direction of the numerical increase in the arithmetic operations should agree with the writing direction because the beginning of the increase (the units) is fixed, while the end of the increase is open. Once again, by knowing that the writing direction in all languages of the Indian subcontinent is from left to right, what then makes the numerical increase to take the contrary direction, if they were the inventors of the decimal positions and therefore the ten digits?
If someone ascribes this Arabic right-to-left direction in the most Indian reading the two-digit numbers to Islam religion as a result of the Islamic conquests to the north of the subcontinent (711 - 1775 AD), this ascribing is refuted by that the influence of the Islamic religion in that early time has been applied to the language and/or script of only those who have embraced Islam, representing what is today Afghanistan and Pakistan, beside Kashmir province. It has not been applied to the script or the language of those who embraced not it, like those who have not got in touch with Arabs of Tamil and remained non-Moslems.
The right-to-left direction of the numerical increase can never be under the Islamic religion's influence in India because this direction has never happened to change to the Indians in the arithmetic operations since they learnt the numerals centuries before Islam. This means that when Indians first learnt to write the numerals then the zero, they have learnt them from the Arabs, the right-to-left direction's proprietors. It is worth understanding that the right-to-left direction of the numerical increase in the arithmetic operations associates not with the reading direction of the finally rendered number to any nation. The same that has not prevented Arabs themselves (after been occupied by the Europeans) to read the greater part of a number first, from left to right, in contrary to the original rule of reading numbers in classic Arabic.
On another side, the four Dravidian languages: Kannada, Telugu, Malayalam and Tamil which are the southern Indian languages and spoken by 27% of the Indian population, and their vast majority are non-Moslems, they read not the two-digit numbers from right to left, and the explanation to this is the early European influence starting with the Portuguese invasion to some of these nations - beginning by 1500 AD as it has passed before us - mainly to the lands of Kannada and Malayalam.
In addition to another partial factor which is the isolation of a nation such as non-Moslem Tamil from the Arabic influence by the dense equatorial jungles encircling their lands, and for this reason too the Arabic zero has not reached the non-Moslem Tamil in the farthest southern of India eastward, while it reached the Dhivehi nation in the farther Maldives who have been reading the two-digit numbers from right - units then tens - but they turned away from this order less than a century past.
On the left, hereabove, I display an analysis for the general transformation I see it arose to the ancient Arabic numerals - following its transmission to India - in writing the modern Devanagari numerals.
