“M is P, S is M, S is P,” the method of my reasoning was almost always in this form:-- “S1 is more or less P, S2 is very similar to S1, S2 is very probably but not certainly more or less P. Let us go on that assumption and see how it works.” That is to say, I was constantly reasoning by analogy and applying verification. So far from using the syllogistic form confidently, I habitually distrusted it as anything more than a test of consistency in statement. But I found the textbooks of logic disposed to ignore my customary method of reasoning altogether or to recognise it only where S1 and S2 could be lumped together under a common name. Then they put it something after this form as Induction:-- “S1, S2, S3, and S4 are P S1 + S2 + S3 + S4 +... are all S All S is P.” I looked into the laws of thought and into the postulates upon which the syllogistic logic is based, and it slowly became clear to me that from my point of view, the point of view of one who seeks truth and reality, logic assumed a belief in the objective reality of classification of which my studies in biology and mineralogy had largely disabused me. Logic, it seemed to me, had taken a common innate error of the mind and had emphasised it in order to develop a system of reasoning that should be exact in its processes. I turned my attention to the examination of that. For in common with the general run of men I had supposed that logic professed to supply a trustworthy science and method for the investigation and expression of reality. A mind nourished on anatomical study is of course permeated with the suggestion of the vagueness and instability of biological species. A biological species is quite obviously a great number of unique individuals which is separable from other biological species only by the fact that an enormous number of other linking individuals are inaccessible in time--are in other words dead and gone--and each new individual in that species does, in the distinction of its own individuality, break away in however infinitesimal degree from the previous average properties of the species. There is no property of any species, even the properties that constitute the specific definition, that is not a matter of more or less. If, for example, as species be distinguished by a single large red spot on the back, you will find if you go over a great number of specimens that red spot shrinking here to nothing, expanding there to a more general redness, weakening to pink, deepening to russet and brown, shading into crimson, and so on and so on. And this is true not only of biological species. It is true of the mineral specimens constituting a mineral species, and I remember as a constant refrain in the lectures of Professor Judd upon rock classification, the words, “they pass into one another by insensible gradations.” It is true, I hold, of all things.