That is always a possibility. One or the other IS in error, either Simpson's Compendium, or the National Archives muster roll card. The question is which one?

Which one do you put your faith and credit in? How can we determine which record is incorrect? Harold Simpson was a noted researcher, but did he miss something? After all he had to be working from the records in the National Archives, which would be the original source of such records, which he would have to compiled his compendium from for the records of the 3rd Arkansas.

But THEN on the other hand, even Bryan H. missed the same card in his roster of the 3rd Arkansas. So we have not One but Two very good noted researchers, who are very well respected, who confirm each other and say that Giles Taylor service ended after June 30, 1863.

So therefore the index card from the National Archives that shows he served in July and August 1863 must be a scribbners error on the part of the transcribers who compiled the CSR way back when. But I have been assured that those people were very careful and double checked their work many time to ensure accuracy. How do you determine that the Transcriber made an error?

It is a Mobius Loop that feeds back upon itself to defy an explaination.