Another suggestion - this is the sort of mistake OCR (optical character recognition) makes. If you try to do a search inside places like Google Books, or anywhere there's digitisation from printed text (older works where the original isn't digital, dodgy ebooks, etc), you'll see a lot of these kind of mistakes. It could just mean whatever you are reading wasn't proofread after OCR, or that it was missed because the person checking just skimmed it and presumed it was ム because 厶 would make no sense.
If that is the issue, you are likely to find similar problems throughout the text.