If you search BMD for the rare surname Jarvin, all records, you get a very odd result. You get very few hits except for deaths in September 1880 when you get a huge number with names starting A to M, and births in June 1910 when you get a huge number of hits with names starting L to W. When you call up the images you can see someone has mistaken the common name JARVIS for the rare name JARVIN. So there are 83 incorrect entries.
Although they offer a mechanism to correct a single entry, it is not clear (to me) how you would try and get a bulk correction like this.
Given Jarvin is so extraordinarily rare I am thinking the few records (apart from the ones I mention above) are due to transcription errors into the original BMD registers. After all, when you are down to just a few individuals you can spot the births, deaths and marriages don't tie up - except for the first 1842 entry of birth and death in Halstead, but that was an area with another unusual surname - Tarvin. So again I suspect whoever copied into the register misread the name.
Steve
