The exchange over that Latin illustrates very clearly the danger of relying on cognates when translating. Had someone been translating from Latin to, say Swahili, and translated mathematicae to whatever the Swahili word is for mathematics, he would have to be considered to have made an error in translation. I am reminded of a rather odd mistranslation in the English translation of GAuss' Disquisitiones in which Gauss is said to have claimed that previous mathematicians have proved it (quadratic reciprocity, no less) by induction but he (Gauss) was giving the first demonstration by infinite descent. Of course, it was a mistranslation to translate probare by prove and inductio by induction (although that is partly because mathematical induction is a rather specialized use of the word even in English). Of course, Gauss was claiming that others had tried it on a lot of numbers, while he was giving the first correct proof, by that form of mathematical induction called infinite descent. Michael