Here is my attempt to write code in CentOS that will read magnetic field data off the summary tapes. I have so far not been successful, even though I have had success with reading plasma data off the summaries (see Case 5 above). If I look in documentation_v_01/Summary_tape_format.pdf on page 1.15-4 I see the variables that should appear in the answer array for the magnetic field after the first 32 words in the header. Those 32 words are described on page 1.15-3. I have just taken all the words in the magnetic field answer array and converted them assuming they are real*4, as I did for the plasma data (see Case 5). This is 568 words or 2272 bytes, and according to the documentation the vast majority of these words are real*4 and should be magnetic field values for the different components and different averaging periods. But when I look at /software_v_01/summary_mag_read/ouput, which lists all the mag answer arrays on day 12 of 1978, I see nothing that looks like what I expect, although this is essentially what I did when I was decoding the plasma data. Very rarely I see something that makes sense, for example word 12 of the header, which is supposed to be “Time period of the data block in seconds”: 48 seconds is the value I expect. But everywhere else the values make no sense.