I suspect that the answer to this puzzle may be something running in parallel to what is known for the 1848 half crown. The article in the 1958 BNJ (p.191-193) using information supplied by Stride refers but the gist is as follows. At the time it was thought that all 1848 2/6ds were 1848/6, but in 1957 two 1848's appeared which were unambiguously not over anything. (The 1848/7 was unknown at this time.) It was known that there were 4 obverse matrices from 1845-1853 and the number of punches was increased from 6-7 in July 1846.Based on the numbers of known dies that were in stock according to the mint records, it was possible to assign the numbers of variously dated dies at the year ends. Year end 1846 there were 35 obv. dies, all of which would have been dated 1846, 28 were sunk between July and December. Jan & Feb 1847 16 obverse dies were sunk but it is not known if they were dated or not (thought unlikely at the time of the article). During the year, 22 of the 35 1846 dated obverses were destroyed. It is thought the remaining 13 were used to strike 1847 coins but without the date altered. 347,488 2/6ds were struck in 1847, though none dated 1847 are known. Only 91,872 were struck in 1848 with the bulk being supplied by recut 1846 dies. At this point in time there was a sharp fall in demand for 2/6ds, so the 13 remaining 1846 dies would have been sufficient to strike the small issue at just over 33000 per die if all were used. All of the 1848 dies examined appeared to have a misplaced E in DEI, but the plain date 1848 is from a die without the misplaced E which ought therefore be one of the 16 dies sunk in 1847 from the new punch, but completed in 1848. The matter is then complicated somewhat by the 1848/7 die which also has a misplaced E in DEI, as do some but not all 1849 and 1850 dies, so the conclusions of the above article need to be revised. For about 5 years from 1847-1851 there was a considerable reduction in mint output with halfcrowns, shillings, sixpences and even small silver having gaps in the dates or very low mintages - probably related to the proposed decimalisation given the large output of Godless florins. Unfortunately, Hocking is not clear on this point as the only things listed are : two halfcrown punches dated 18-- (item nos. 1263 & 1264), plain fillets, one with w.w. on the truncation and also a matrix (item 1250) as for the 1839 proof but with 18--. It is therefore unclear where Stride got his information from. However, the existence of an 1847 sixpence is quite feasible based on the inferences that could be drawn from the half crowns. All that would be needed would be an undated die or one dated 18--. This would be in keeping with a misaligned date and from known pennies etc and retained mint material we know that dies were kept with either none or a partial date.