These sounds were unkown in Latin, as far as I know. For my Latin translation of Vathek, there are some Arabic (or Arabic-looking) proper names that contain those sounds (e.g. Edris Al Shafei, Gulchenrouz, Shaddukian) and I'm unsure what to do with them. Thus far I've kept them all as they are and just stuck Latin endings at the end of most of them, but this may not be the best way to go. In one 19th-century book on Google Books I've found the adjective Casmirensis which presumably means "of Kashmir" (which was a word I needed at some point in my translation). Previously, I had seen "chocolate" rendered in Neo-Latin as socolata. Here, admittedly, I can't tell with certainty whether this "s" was meant to stand for the "sh" sound that we've got there in French or for the "ch" sound we've got in English and certainly some other languages (even Romance ones; I've just had a listen to the Spanish word and it sounds like English "ch" in the beginning). Personally, however, a slightly more logical representation of the "ch" sound would seem to me to be ts — but was it done this way in, say, medieval or Renaissance Latin? I generally like to base my choices on older attestations when possible. So does anyone know whether there was any customary way of representing those sounds in Latin at some point?