v's and r's were difficult to distinguish between and so were u's and n's in some instances. What looked like superscript >'s could actually have been a well-formed apostrophe. Quality of the photocopy of the original script wasn't always clear, and there may be some accents over the e's that didn't show up and therefore I haven't recorded them.
Andreea: replaced /đ/ with /ð/ - not 100% of this, but following Ferrell et. al.'s conventions, it should be correct. (Clearly the /ð/ symbol has not be used elsewhere in the data.) (8 changes in total)