In Ancient Quenya, labial [ms] became [ns], and later this [ns] became [ss]. The fullest description of this intermediate sound change appears in the Outline of Phonology from the 1950s (OP2):

[Nasals] before s. In CE [Common Eldarin] the assimilations in this case precede as before voiceless stops. At the end of consonant-groups s was too strong to be voiced, while it was not strong enough to unvoice the nasal. Thus in CE as nt, ñt > nt, so ns, ñs > ns, but mt, ms remained. In Quenya ms was assimilated > ns (as mt > nt). The resultant ns, from any nasal + s, was then in the oldest strata unvoiced to ss (OP2, PE19/98-99).

This sound change was discussed elsewhere among Tolkien’s statements that all nasals + s became [ss], a phonetic development he noted in both Outline of Phonetic Development from the 1940s (OP1) and from the Outline of Phonology from the 1950s (OP2):

ñs, ms in old formations similarly > ns. Since nasals [n] were unvoiced before [s], the phonetic product was [ss] of all nasals + [s] (OP1, PE19/47).

ns from n/m/ŋ. This soon became ss (with unvoicing of n) (OP2, PE19/82).

As indicated in the very first quote above, the sound change whereby velar [ŋs] became [ns] occurred in Primitive Elvish and was thus common to all Eldarin languages. This primitive velar assimilation was also noted in the second version of Tengwesta Qenderinwa (TQ2) from the 1950s:

The following general changes belong to CQ [Common Quenderin] or are at least anterior to PE [Primitive Eldarin] ... The nasal [ŋ] became n before t, th, d, s (TQ2: PE18/101-102).

Thus ms > ns was a later and possibly Quenya-only change. There is no sign of this sound change in Sindarin or Noldorin, but there are no counter-examples either: the combination ms does not appear in any published Eldarin words from the 1930s and later.

Conceptual Development: There are not enough examples to determine whether this sound change was present in the Early Qenya of the 1910s and 1920s, but given the general pattern of nasal assimilations in this early conceptual stage, it seems likely it would have occurred. Our only concrete example for the development of this primitive sound combination in the Early Period is for a Gnomish word:

• ᴱ✶amsŏrı̯a > G. annuir “lofty” (PE13/110).

Here it seems that perhaps [ms] > [ns] > [nn]. If so, then [ms] > [ns] may have been a phonetic development in Primitive Elvish in the Early Conceptual Period.