Among the swords of mainland Southeast Asia that are circulating between dealers and collectors is a group of old swords in excavated condition. Styles within this group vary, and they appear to represent different time periods and cultures.

Little is known about them. Some of them are identified as from the Khmer empire of  802-1431 AD, while a group of swords in the History Museum, Hue, Central Vietnam are dated to the Tây Sơn dynasty of 1778–1802 AD.

These pieces are of generally low value, too low to justify thorough research from a commercial perspective, but sometimes I need to let passion and curiosity reign. It's why I got into this business in the first place! So when I had two in my possession I decided to run some scientific analysis on them to see if we can find out more.


The methods; X-ray fluorescence & radiocarbon dating

The research conducted consisted of X-ray fluorescence (XRF) and radiocarbon dating of fragments of wood.

XRF is a non-destructive method to determine the composition of a material by exposing it to high-energy gamma rays, after which it briefly emits fluorescent X-rays peculiar to each element. Older materials are not as homogenous as modern materials, so you can expect to get a slightly different reading at each spot of a single piece. Also, it tests the surface which may give different results from the core. Silicon for example typically binds to the metal when pieces are in soil for some time, while not originally present in the alloy. Elements lighter than magnesium cannot be measured using XRF, so unfortunately it cannot detect the levels of carbon in the iron.

Radiocarbon dating determines the amount of radioactive carbon isotope 14C there is in a material. 14C is constantly created in the earth's atmosphere and absorbed by plants through photosynthesis. When the plant dies, or the animal that eats the plant stops consuming, the level of 14C in it will gradually decline. 14C levels were generally relatively stable, making fairly accurate dating possible for most periods. From the 18th century onwards, atmospheric and human activity distorted the 14C content in the atmosphere, giving a much wider range of possible dates for later pieces.

I will present the data pretty much as I got it, hoping it will serve the researching and collecting community. I'd be happy to learn about any observations a reader may make from them.


1. Mainland Southeast Asian sword with eight-lobed guard

Excavated mainland Southeast Asian swordLobed guard of southeast Asian sword


XRF results

Hilt guard side (brass)
72.7% copper
22.5% zinc
2.6% silicon
Trace elements below 1% of antimony, tin, lead, nickel, iron.

Hilt pommel side (leaded copper alloy)
72.3% copper
6.6% lead
4.5% zinc
4% aluminum
2.7% tin
8.4% silicon
Trace elements below 1% of antimony, nickel, iron, titanium, bismuth, phosphorous.

Pommel cap (copper alloy)
77.7% copper
5.1% tin
3.1% aluminum
1.7% lead
1% phosphorous
10% silicon
Trace elements below 1% of antimony, bismuth, gold, nickel, iron, titanium, manganese.

Guard (leaded copper alloy)
72.5% copper
7.1% lead
5.8% tin
4.3% aluminum
2.8% zinc
1.2% iron
5.9% silicon
Trace elements below 1% of antimony, zirconium, bismuth, nickel, manganese.

94.1% iron
1.8% aluminum
2.3% silicon
Trace elements below 1% of molybdenum, tungsten, copper, nickel, sulfur.

It is interesting to see how much the composition of the alloys differs per piece, indicating perhaps a production line where different parts came from different batches or sources. The somewhat erratic compositions are also indicative of traditional manufacture. Modern materials tend to be more homogenous in composition.

The high levels of silicon are usually indicators of a piece that has been in the ground for some time.


Radiocarbon dating

A piece of old wood deep into the guard side hilt sleeve was prepared for carbon dating. It gave the following results:

68.3% probability

1671 AD - 1695 AD (11.4%)
1725 AD - 1779 AD (25.3%)
1798 AD - 1812 AD ( 6.6%)
1839 AD - 1845 AD ( 2.2%)
1852 AD - 1877 AD ( 8.7%)
1916 AD - 1944 AD (14.1%)

95.4% probability

1665 AD - 1711 AD (16.5%)
1717 AD - 1785 AD ((29.1%)
1794 AD - 1823 AD  (10.1%)
1831 AD - 1894 AD (20.6%)
1905 AD - ... (19.2%)

(OxCal v4.4.2 Bronk Ramsey, 2020. Atmospheric data from Reimer et al, 2020.)


The results are quite broad, due to the abovementioned 14C instabilities from the 18th century onwards. The piece most likely dates from the 18th century, but with possibilities of being late 17th century to the early 20th century.

There is one caveat, the dating is based on a sliver of wood that was lodged inside the hilt. Hilts break and their wood can be replaced at later times.


2. Mainland Southeast Asian dha sword with scabbard

Old dha with scabbard


XRF results

Scabbard plating (brass)
60.1% copper
22.1% zinc
5.4% aluminum
1.4 % lead
9.1% silicon.
Trace elements below 1% of iron, zirconium, nickel, tin, phosphorus.

60.6% copper (brass)
24.1% zinc
4.6% aluminum
2.3% lead
1.2% tin
6.6% silicon
Trace elements below 1% of iron, zirconium, nickel, phosphorus.

50% copper (brass)
18% zinc
10.1% aluminum
1.3% tin
1.2% iron
14.9% silicon
Trace elements below 1% of zirconium, nickel, bismuth, phosphorus, antimony, titanium.

94.1% iron
2.3% silicon
Trace elements below 1% of molybdenum, tungsten, copper, nickel, sulfur.

The high levels of silicon are usually indicators of a piece that has been in the ground for some time.


Radiocarbon dating

A small hole was drilled in the hilt of this sword, pommel side, and wood was extracted from it. We made sure only core wood was taken, the material near the exposed surface was discarded. It gave the following results:

68.3% probability:

1640 AD - 1672 AD (42.7%)
1778 AD - 1799 AD (25.6%)

95.4% probability:

1528 AD - 1550 AD (3.5%)
1634 AD - 1684 AD (48.4%)
1735 AD - 1804 AD (39.7%)

(OxCal v4.4.2 Bronk Ramsey, 2020. Atmospheric data from Reimer et al, 2020.)

From the above, we can conclude that the piece is most likely from the 17th or 18th century, with a very slight chance it is from the early 16th century.


Probably origins

In conversations with collector Iain Norman he pointed out how many of these in local museums seem to be attributed to the Battle of Rạch Gầm-Xoài Mút. It was fought between the Vietnamese Tây Sơn forces and an army of Siam in present-day Tiền Giang Province in the Meikong river delta in the south of Vietnam on January 20, 1785.

On the Siamese side there were some 20.000 sailors and 30.000 infantry, using 300 warships, aided by some 3000-4000 troops of the north Vietnamese Gia Long emperor. On the Tây Sơn side were 30.000 men with 55 warships, 100 sailing boats and 300 canoes.

It resulted in a decisive Tây Sơn victory, with an estimated 40.000 killed on the Siamese side, and their entire navy destroyed. Today, treasure hunters still dredge and dig up swords from this battle, and it seems very likely that the ones on the market -including these- can be dated back to that event.

Do you have anything for sale?

I might be interested in buying it.

Contact me

Found in excavated condition, published with results of c-14 and XRF analysis.


In excavated condition. With XRF and radiocarbon dating results.


A very rare ceremonial variety with copper scabbard inlaid with different alloys, and a brass blade.


Very rare set of Burmese knives from Mindan village.


For holding and protecting important documents.


An exceedingly rare set with fine mother of pearl inlaid string board