Here 𝓟 lacks the prepositional a before marching, while the 1830 edition has the a. In the 1837 edition, all remaining instances of this dialectal use of a were permanently removed from the text. As explained under Helaman 16:4, there were originally 26 cases of prepositional a in the copytext for the 1830 typesetter, who omitted the a in 11 of those cases. But there is no independent evidence that he himself ever added the prepositional a. We should consider Mormon 6:7 one of those cases where he maintained the prepositional a (for a total of 15 times); it is rather unlikely that he would have added the a here since he never did elsewhere.
Unfortunately, we have no other examples showing scribe 2 of 𝓟 either omitting or adding the prepositional a. There are three instances of the prepositional a in his hand, all in Alma 10; each of these was written without correction. But when we consider Oliver Cowdery’s practice, we find that in four cases he momentarily omitted the a; and in only one case (marked below with an asterisk) did he accidentally add the a, again momentarily:
There are two cases, both in 𝓞, where Oliver consciously removed the prepositional a. One is listed above (in the first instance of the prepositional a in Alma 28:5); here is the second one:
(See under Alma 28:5 and Alma 55:8 for discussion of these two conscious deletions of the a.) All in all, these statistics suggest that sometimes the scribe omitted the prepositional a, although not as frequently as did the 1830 typesetter. Here in Mormon 6:7, the critical text will accept the 1830 reading with the prepositional a as the probable reading in 𝓞, which means that scribe 2 of 𝓟 omitted it when he copied the text from 𝓞 into 𝓟.
Summary: Restore in Mormon 6:7 the prepositional a of the 1830 edition: “my people ... did now behold the armies of the Lamanites a marching towards them”; 𝓟 lacks the a here, but this is likely the result of scribe 2 of 𝓟 omitting the a since the 1830 typesetter never otherwise added the prepositional a.