How did they measure sword blades?

[Jens Nordlunde]

In The Wonder that was India (p. 503) 1 angula is 8 barly corns or 3/4 in.

In The Jahangirnama (p. 473) a fingers width is said to be 2.032 cm 4/5 in..

In Hidu Arms and Ritual (p. 205) the best sword blades should be 100 fingers (101.6 cm) long, a nimcha has a blade of 50 fingers (50.8 cm), and other blades have a length between these two measurers. It is, however, mentioned that a sword blade should be an even number of fingers, or all kinds of misfortune will fall upond you head.

But how did they measure the blades? From the hilt to the tip in a straight line, or along the back of the blade?

I have tried to do both on two of my blades.
The Indian blade is 77 cm long measured in a straight line, and only a fraction longer measured along the back of the blade. This gives 37.893 fingers, so I would say it is 38 fingers - which is an even number.

The shamshir blade is 79 cm long mesured in a straight line, or 38.879 fingers, so I would say 39 fingers - an uneven number.
But measured along the back of the blade it is 81 cm or 39.862 fingers, so I would say 40 fingers - and even number.

In catalogues I think measurers are mostly given measured in a straight line, but from the measurers above this may not be the way the did it in India in the old days.

Any comments will be welcome.