was forced to suppose that the calcination of mercury in the air must be a more complex occurrence than merely the expulsion of phlogiston from the mercury:  for, if the process consisted only in the expulsion of phlogiston, how could heating what remained produce exceedingly pure ordinary air?  It seemed necessary to suppose that not only was phlogiston expelled from mercury during calcination, but that the mercury also imbibed some portion, and that the purest portion, of the surrounding air.  Priestley did not, however, go so far as this; he was content to suppose that in some way, which he did not explain, the process of calcination resulted in the loss of phlogiston by the mercury, and the gain, by the dephlogisticated mercury, of the property of yielding exceedingly pure or dephlogisticated air when it was heated very strongly.

Priestley thought of properties in much the same way as the alchemists thought of them, as wrappings, or coverings of an essential something, from which they can be removed and around which they can again be placed.  The protean principle of phlogiston was always at hand, and, by skilful management, was ready to adapt itself to any facts.  Before the phenomena of combustion could be described accurately, it was necessary to do two things; to ignore the theory of phlogiston, and to weigh and measure all the substances which take part in some selected processes of burning.

Looking back at the attempts made in the past to describe natural events, we are often inclined to exclaim, “Why did investigators bind themselves with the cords of absurd theories; why did they always wear blinkers; why did they look at nature through the distorting mists rising from their own imaginations?” We are too ready to forget the tremendous difficulties which beset the path of him who is seeking accurate knowledge.

  “To climb steep hills requires slow pace at first.”

Forgetting that the statements wherein the men of science of our own time describe the relations between natural events are, and must be, expressed in terms of some general conception, some theory, of these relations; forgetting that the simplest natural occurrence is so complicated that our powers of description are incapable of expressing it completely and accurately; forgetting the uselessness of disconnected facts; we are inclined to overestimate the importance of our own views of nature’s ways, and to underestimate the usefulness of the views of our predecessors.  Moreover, as naturalists have not been obliged, in recent times, to make a complete renunciation of any comprehensive theory wherein they had lived and moved for many years, we forget the difficulties of breaking loose from a way of looking at natural events which has become almost as real as the events themselves, of abandoning a language which has expressed the most vividly realised conceptions of generations of investigators, of forming a completely new mental picture of natural occurrences, and developing a completely new language for the expression of those conceptions and these occurrences.