This query gives 12 hits because, for example, CSH1_HUMAN and CSH2_HUMAN differ by
only one amino acid. Probably they were one sequence with a variant till it was
realized that for one biological reason or the other (like there really being two genes or so)
these actually are two sequences. See also the eduwiki entry for accession code.
SOMA_CAPHI has a secondary accession code that starts accidentally with P0124. If you search for ac:P01247, you get two hits (one from a sheep and one from a goat). Now things look like extremely confusing. Didn't we tell you that the accession codes were unique?
You kids are too young to know that there was a time that it was difficult to store the SwissProt database on a big computer. But, believe me, your mobile phone has more compute power than the super computer I used to work on for my PhD... So, one day, the people of SwissProt decided that if two sequences were entirely identical, they would merge the files. Tho keep the accession codes unique they simply gave a merged file both accession codes, but further they normally needed not much extra. they must mention two species, but the function of the protein is, of course, the same, etc.
In the late 90′s, computers started to grow faster than SwissProt, and then it was decided to split the sequences again. Arghh... So now the split files got their original (pre-merge) accession code back as primary accession code, and the extra accession code it obtained upon merging now became a secondary accession code.
This is just one scenario. There are other complicated things that can happen with data. And consequently some files have many accession codes. But... If you find an accession code in the literature, and you find the corresponding SwissProt entry with MRS, then the two DO belong together!
See http://www.uniprot.org/help/accession_numbers if you want to know more about this stupid topic (the eduwiki accession code page also holds a pointer to the UniProt page.