Most likely, there are four marriages on the page and the page number for the missing fourth male has been mistranscribed. Second most likely, there are only three marriages on the page, and the page number for the extra fourth female has been mistranscribed. I have resolved this sort of thing in the past by downloading an entire quarter's marriages for the Exeter registration district, putting them into Excel to make them easier to manipulate, and then finding (a) the entries with obviously garbled page numbers and (b) the entries on pages with a mismatched number of males and females. Then look at the images for these entries to work out what has gone wrong in each case. If you want to try this, and my explanation is unclear, let me know. If you do find a transcription error (or more than one) then be sure to submit a correction to FreeBMD.

The entry for William SEAWARD (Exeter 10 134) has been correctly transcribed from the typed index, but 134 is out of the expected range for Exeter registration district. I would place a small bet that William's correct reference is page 184, and that the typist who produced the typed index from the original handwritten index misread an 8 as a 3.