Re: Sound changes causing divergence of ordinals from cardinals

Another case, perhaps, of ordinals that are not regularly relatable
(anymore) to the cardinal, my Gwr lang., which is monosyllabic and tonal.
The proto-lang. was mainly CVCV(C), and one could form adjectives with a
prefix *m-

one: ay (hi) < had, first fay (mid) < m+had
two: ni (hi) < hVní, 2nd di (mid) < m+(hV)ni
three: hr (lr) < hVqV[+str]r, 3rd gr, Nr or kr (undecided)
(four exists, but for some reason isn't in the dictionary yet...)
five; fifth: dzi (low) < dVsí
six: ho (hi) < qahú,  sixth go, No, or ko
seven: shr (hf) *sí(dg)Vr, jr (m) seventh
(eight: fang (m), 8th pang (m) < Kash fanu)
(nine: sang (m), 9th tsang (m) < Kash sana)
ten (decimal, old octal): maq (h) < *kVmáC; baq (m) tenth

hundred (decimal) : loq (m), hundreth doq (m) < **rog < Kash roNgo
thousand (decimal, old octal) : chih (h) < *kitís,  giq (m) thousandth (app.
< m+kítis

One might get the impression that not all sound changes have been worked
out....A project for dreary winter days.