Just as a follow-up on Mikes observations about "numbering", the history of GTh's segregation into saying and sub-sayings is somewhat interesting (to me at least). As I mentioned in a previous post, Pahor Labib's numbering scheme was the first attempt I know of to segregate the text for the purpose of reference. It used a combination of "Plate Number" plus the line number and completely ignored anything to do with "context". The plate number was actually the photograph number so that, for example what we conventionally call GTh1 would have been cited as 80.14-19. Hence the first printed edition of the Gospel of Thomas that I purchased in 1969 still retained Labib's schema as well a set of parenthetic divisions proposed by Guillaumont, Puech, Quispel, Till, al Masih and Quecke (who were incidentally editors of the copy I bought back then). The latter divided the text into 114 discrete sections and had become the accepted convention by the time the Brill critical edition was published in 1989.
There were at least three other numbering schemes that preceded the one I mention above. The trhee differed not only in the number of sayings (112, 113 and 113) but also at the location in the text where the divisions were to be placed. Just for the sake of example, what we know these days as GTh 59 was designated 60, 58 or 64 depending on which scholar's work was being studied.
Then of course there is the edition of the Gospel of Thomas rendered into Greek with a French translation and commentary by Kasser that appeared in 1961. The "retoverted" text (Coptic into Greek) was not the only unique feature of the book; it also divided Thomas into 250 "versets". By comparison the most recent division segregates the text into 315 or 316 units (if each sub saying is counted as a separate division).