Experimental thinking, or scientific reasoning, is thus a conjoint process of _analysis and synthesis_, or, in less technical language, of discrimination and assimilation or identification. The gross fact of water rising when the suction valve is worked is resolved or discriminated into a number of independent variables, some of which had never before been observed or even thought of in connection with the fact. One of these facts, the weight of the atmosphere, is then selectively seized upon as the key to the entire phenomenon. This disentangling constitutes _analysis_. But atmosphere and its pressure or weight is a fact not confined to this single instance. It is a fact familiar or at least discoverable as operative in a great number of other events. In fixing upon this imperceptible and minute fact as the essence or key to the elevation of water by the pump, the pump-fact has thus been assimilated to a whole group of ordinary facts from which it was previously isolated. This assimilation constitutes _synthesis_. Moreover, the fact of atmospheric pressure is itself a case of one of the commonest of all facts--weight or gravitational force. Conclusions that apply to the common fact of weight are thus transferable to the consideration and interpretation of the _relatively_ rare and exceptional case of the suction of water. The suction pump is seen to be a case of the same kind or sort as the siphon, the barometer, the rising of the balloon, and a multitude of other things with which at first sight it has no connection at all. This is another instance of the synthetic or assimilative phase of scientific thinking. If we revert to the advantages of scientific over empirical thinking, we find that we now have the clue to them. [Sidenote: Lessened liability to error] (_a_) The increased security, the added factor of certainty or proof, is due to the substitution of the _detailed and specific fact_ of atmospheric pressure for the gross and total and relatively miscellaneous fact of suction. The latter is complex, and its complexity is due to many unknown and unspecified factors; hence, any statement about it is more or less random, and likely to be defeated by any unforeseen variation of circumstances. _Comparatively_, at least, the minute and detailed fact of air pressure is a measurable and definite fact--one that can be picked out and managed with assurance. [Sidenote: Ability to manage the new] (_b_) As analysis accounts for the added certainty, so synthesis accounts for ability to cope with the novel and variable. Weight is a much commoner fact than atmospheric weight, and this in turn is a much commoner fact than the workings of the suction pump. To be able to substitute the common and frequent fact for that which is relatively rare and peculiar is to reduce the seemingly novel and exceptional to cases of a general and familiar principle, and thus to bring them under control for interpretation and prediction.