There are two things, then, which it is more especially desirable to keep constantly before one in reading Balzac--two things which, taken together, constitute his almost unique value, and two things which not a few critics have failed to take together in him, being under the impression that the one excludes the other, and that to admit the other is tantamount to a denial of the one. These two things are, first, an immense attention to detail, sometimes observed, sometimes invented or imagined; and secondly; a faculty of regarding these details through a mental lens or arrangement of lenses almost peculiar to himself, which at once combines, enlarges, and invests them with a peculiar magical halo or mirage. The two thousand personages of the _Comedie Humaine_ are, for the most part, "signaled," as the French official word has it, marked and denoted by the minutest traits of character, gesture, gait, clothing, abode, what not; the transactions recorded are very often given with a scrupulous and microscopic accuracy of reporting which no detective could outdo. Defoe is not more circumstantial in detail of fact than Balzac; Richardson is hardly more prodigal of character-stroke. Yet a very large proportion of these characters, of these circumstances, are evidently things invented or imagined, not observed. And in addition to this the artist's magic glass, his Balzacian speculum, if we may so say (for none else has ever had it), transforms even the most rigid observation into something flickering and fanciful, the outline as of shadows on the wall, not the precise contour of etching or of the camera. It is curious, but not unexampled, that both Balzac himself when he struggled in argument with his critics and those of his partisans who have been most zealously devoted to him, have usually tried to exalt the first and less remarkable of these gifts over the second and infinitely more remarkable. Balzac protested strenuously against the use of the word "gigantesque" in reference to his work; and of course it is susceptible of an unhandsome innuendo. But if we leave that innuendo aside, if we adopt the sane reflection that "gigantesque" does not exceed "gigantic," or assert as constant failure of greatness, but only indicates that the magnifying process is carried on with a certain indiscriminateness, we shall find none, I think, which so thoroughly well describes him.