No, the book says "pertistent", and the transcriber mistakenly wrote "persistent". Neither of those is a recognition error. Rather, the error is that the transcriber did recognize what was meant, and ignored what was actually written.
> If another copy from the same edition has the error corrected, such cues may help to identify early and late printings and contribute to a more comprehensive account of the book’s printing history.
In other words, when transcribing books you want to preserve misspellings that occur in the source text.
It’s actually quite interesting because that means that automatic spellchecking of OCRed text while helping to improve the quality of the transcript could also introduce unwanted corrections. But doing like the OP did and comparing their transcripts with those of Google Books was clever.
Tangent: scribal errors are often classified as to whether they are committed by scribes who do understand the language they're copying or scribes who don't. (Some errors can be committed by either kind of scribe, but will still tend to lean one way or the other.) Copying "persistent" where the text has "pertistent" is a good example of a kind of error that only a scribe who understands the text will make. (Though this particular case might not even be considered a scribal error.)
I was confused by your reference to understanding the text, so I looked it up. I think a scribe who really understood the text would recognise the intentional misspelling. The transcription error reflects a lack of understanding.
What intentional misspelling? "Pertistent" in the 1879 printing is a printer's error, and it's very clear if you look at the passage that it can't be intentional, because the same character uses the word 5 times in quick succession.
The transcriber's error is unwitting; he specifically comments on the fact that he didn't want to make it.
