As a preliminary, he has a short chapter to show that there is variation in Nature, and therefore something for natural selection to act upon. He readily shows that such mere variations as may be directly referred to physical conditions (like the depauperation of plants in a sterile soil, or their dwarfing as they approach an Alpine summit, the thicker fur of an animal from far northward, etc.), and also those individual differences which we everywhere recognize but do not pretend to account for, are not separable by any assignable line from more strongly-marked varieties; likewise that there is no clear demarkation between the latter and sub-species, or varieties of the highest grade (distinguished from species not by any known inconstancy, but by the supposed lower importance of their characteristics); nor between these and recognized species. "These differences blend into each other in an insensible series, and the series impresses the mind with an idea of an actual passage." This gradation from species downward is well made out. To carry it one step farther upward, our author presents in a strong light the differences which prevail among naturalists as to what forms should be admitted to the rank of species. Some genera (and these in some countries) give rise to far more discrepancy than others; and it is concluded that the large or dominant genera are usually the most variable. In a flora so small as the British, 182 plants, generally reckoned as varieties, have been ranked by some botanists as species. Selecting the British genera which include the most polymorphous forms, it appears that Babington's Flora gives them 251 species, Bentham's only 112, a difference of 139 doubtful forms. These are nearly the extreme views, but they are the views of two most capable and most experienced judges, in respect to one of the best-known floras of the world. The fact is suggestive, that the best-known countries furnish the greatest number of such doubtful cases. Illustrations of this kind may be multiplied to a great extent. They make it plain that, whether species in Nature are aboriginal and definite or not, our practical conclusions about them, as embodied in systematic works, are not facts but judgments, and largely fallible judgments-