The use of but in the printer’s manuscript does not seem appropriate, while the use of and in the 1830 edition does. One could argue, of course, that the difficult reading with but was the reading of the original manuscript and that the 1830 typesetter changed the difficult but to and. Another possibility is that Oliver Cowdery accidentally substituted but for an ampersand when he was copying from 𝓞 into 𝓟. Such a change could have been caused by the visual similarity of but with the beginning of the next word, behold.
We can find evidence that Oliver Cowdery occasionally mixed up and and but; in fact, we have evidence for his making errors in either direction, as in the following two examples:
The first error (in 2 Nephi 15:7) was probably influenced by the preceding use of and in “and behold oppression”. The second example clearly shows that Oliver can accidentally replace an ampersand with but. (For other examples where Oliver mixed up and and but, see the list under 1 Nephi 8:20, in the discussion regarding straight versus strait.) On the other hand, we also have evidence that the 1830 typesetter was occasionally willing to edit out what he considered inappropriate uses of but when and was expected or worked better. As an example, see the (rather complex) discussion under 3 Nephi 4:15–16.
Given this mixed evidence, it is probably best to assume that the original text made sense, which would imply that the original manuscript read “and behold” here in 3 Nephi 19:6. The critical text will assume as much and accept the 1830 reading as the original one.
Summary: Accept the 1830 reading “and behold” in 3 Nephi 19:6 since the original manuscript could well have read this way; in addition, and works much better than but.