to the reader, and, moreover, it may have been the author's starting point. The corresponding passage from Martorell's work reveals that the source was followed very closely. It is an easy matter to pick out the part of it that he needed to change in order to make it suit his purpose. The details of the latter part give evidence that he consulted the source probably more than once, and that the modifications which we find there are not necessarily due to the processes of translation. The same may be said of the second parallel passages. But the other three are so very nearly alike that the translation theory must be given up. Let us remember what we are asked to believe. We are told that Tirant lo Blanch was translated from the English. Then these passages were translated from Lull's work (written in Catalan), first into English, then into Portuguese, and finally back into Catalan. And in all these various processes the passages in their final form are practically the same as the original. Is not this a marvelous achievement? We cannot believe the author's ingenious representations. We are convinced that Tirant lo Blanch was not translated from the English. And since that is the case, is it not probable that it was not written in Portuguese? We admit that by translating from the Catalan into Portuguese, and then back again into Catalan, the resulting passages might not be so very different. But we cannot believe that they could be so similar to the original after two translations. Then, how do we account for the changes? A comparative study of Lull's style and language and that of Martorell shows that there was considerable difference between them. Of course, it was not the latter's intention to copy the material word for word. He recast it, adapting it to his style and vocabulary; in other words, he took in general merely the ideas and expressed them in his own way. If Lull's manner of expression in certain cases agreed with
