facts and invented part of her story; and so it all ends, and they live happily ever after! Feeble and untrue as are plot and characters in this "poor novel" (as Carlyle without injustice called it), yet many scenes are well written, the details are truly colored, and every page is illuminated with thought of so high an order and language so brilliant, so flowing, so felicitous, that one forgives, for the sake of merits such as these, the failure of the fiction to be either true or interesting. This seemed to show, nevertheless, that Harriet could write essays, and travels, and didactic and philosophical works, but could not write a novel except "with a purpose," when the accomplishment of the purpose might excuse any other shortcomings. But when one considers the great excellence of many of the _Illustrations_, the decided drawing of the characters, the truthful analysis of the springs of human action, the manner in which the incidents are combined and arranged to develop and display dispositions and histories, it becomes clear that she _had_ great powers as an imaginative depicter of human nature and social life, and that there must have been other causes than sheer incapacity for the faults and the feebleness of _Deerbrook_. The first cause was what seems to me a mistaken theory about plots in fiction, which she had adopted since writing the _Illustrations_. She now fancied that a perfect plot must be taken from life, forgetting that we none of us know the whole plot of the existence of any other creature than ourselves, and that the psychological insight of the gifted novelist is displayed in arguing from what is known to what is unknown, and in combining the primary elements of human character into their necessary consequences in act and feeling. This error she would have been cured from by experience had she gone on writing fiction. She might have been aided in this by what she naively enough avows about _Deerbrook_: that she supposed that she took the story of Hope's marriage from the history of a friend of her family, and that she afterwards found out that nothing of the sort had really happened to him! She might then have asked herself whether the story as she had told it was more possible than it was possible that gunpowder should be put to flame without an explosion. A girl in her teens might have been forgiven for playing with the history of the wildest passions of the human heart; but Harriet Martineau erred because she tried to enslave herself to fact in a matter in which she should have inferred, judged from psychological principles, and trusted to the intuitions of her own mind for the final working out of her problem. As it was, if her "fact" had been a reality we should have been compelled to account for the placid progress of events by the supposition that she had utterly misrepresented the characters of the persons involved.