metallic gases, and the formation of a photosphere at that outer limit where the gases condense; accepting the suggestion of Sir John Herschel, so amply supported by evidence, that zones north and south of the Sun's equator are subject to violent cyclones; this hypothesis is, that if a cyclone occurs within the atmosphere of metallic gases between the molten surface and the photosphere, its vortex will become a region of rarefaction, of refrigeration, and therefore of precipitation. There will be formed in it a dense cloud extending far down toward the body of the sun, and obstructing the greater part of the light radiating from below. Here we have an adequate cause for the formation of an opaque vaporous mass—a cause which also accounts for the frequently-observed vortical motion; for the greater blackness of the central part of the umbra; for the formation of a penumbra by the drawing in of the adjacent photosphere; for the elongation of the luminous masses forming the photosphere, and the turning of their longer axes toward the centre of the spot; and for the occasional drifting of them over the spot toward the centre. Still, there is the difficulty that vortical motion is by no means always observable; and it remains to be considered whether its non-visibility in many cases is reconcilable with the hypothesis. At present none of the interpretations can be regarded as established.
Here are sundry suppositions which the man of science severally tests by observations and necessary inferences. In this, as in other cases, he rejects such as unquestionably disagree with unquestionable truths. Continually excluding untenable hypotheses, he waits to decide among the more tenable ones until further evidence discloses further congruities or incongruities. Checking every statement of fact and every conclusion drawn, he keeps his judgment suspended until no anomaly remains unexplained. Not only is he thus careful to shut out all possible error from inadequacy in the number and variety of data, but he is careful to shut out all possible error caused by idiosyncrasy in himself. Though not perhaps in astronomical observations such as those above implied, yet in all astronomical observations where the element of time is important, he makes allowance for the intervals occupied by his nervous actions. To fix the exact moment at which a certain change occurred, his perception of it has to be corrected for the "personal equation." As the speed of the nervous discharge varies, according to the constitution, from thirty to ninety metres per second, and is somewhat greater in summer than in winter; and as, between seeing a change and registering it with the finger, there is an interval which is thus appreciably different in different persons; the particular amount of this error in the particular observer has to be taken into account.
Suppose now, that, to a man of science, thus careful in testing all possible hypotheses and excluding all possible sources of error, we put a sociological question—say, whether some proposed institution will
