ured. Our views did not go beyond that; but we had scarcely made any progress with the book when we found that it was impossible to keep the translation as it stood. Obvious omissions, mistaken renderings of the sense, originating, no doubt, with the faulty Arab text which the translator had at his disposal, and which were patent at first sight, imposed upon the necessity of consulting other sources. We were thus induced to examine all the Arab manuscripts of the work which we could by any possibility obtain.
Three texts were to this end put under contribution. These treated of the same subjects in the same order, and presented the same succession of chapters, corresponding, however, in this respect, point by point, with the manuscript upon which our translator had to work, but while two of them gave a kind of abstract of the questions treated, the third, on the contrary, seemed to enlarge at pleasure upon every subject.
We shall expatiate to some slight extent upon this last named text, since the study of it has enabled us to clear up a certain number of points upon which M———, notwithstanding his conscientious researches, has been unable to throw sufficient light.
The principal characteristic of this text which, is not exempt from gross mistakes, is the affectation of more care as to style and choice of expressions; it enters more into fastidious, and frequently technical particulars, contains more quotations of verses—often, be it remarked, inapplicable ones—and uses, in certain circumstances, filthy images, which seem to have had a particular attraction for the author; but as a compensation for these faults, it gives, instead of cold, dry explications, pictures
