It is sometimes dangerous reasoning from a metaphor. It frequently presents but one aspect of a truth, and the changing or inverting that aspect may invert the whole argument built upon it. It is very common, for example, to compare knowledge to heat. We lately read what the speaker doubtless regarded as a very imposing argument, grounded wholly upon such a simile. He was contending, with the greatest moral courage, that our common schools should receive the most liberal patronage of the State, while the colleges should be "left to themselves." "Knowledge," says the undaunted advocate of this very unpopular doctrine, "knowledge will no more descend than heat will descend. If you wished to warm the lower stratum of air, would you heat the upper stratum first? No, sir! Warm the lower stratum, and then you can not keep the upper cold." We know not which to admire most here, the science or the logic. A pretty good argument in favor of a higher education for legislators might be deduced from it, but not in such a way, perhaps, as the orator imagined. Knowledge then is heat. Heat ascends. Ergo, the common schools are the foundation and, therefore, keeping the stove well supplied below is certainly the best means of warming the dummy above. Admirably argued. But let us now change the metaphor. Knowledge is _light_. This must strike most minds as being, to say the least, quite as appropriate a simile as the other. Knowledge is light, and light comes down. Its native seat is in the upper region. Where now is our metaphorical argument? Turned upside down, and every inference pointed like a battery against the very positions it was intended to support. With the change of a very few terms all that follows becomes a parody on the former meaning. "If you wish to _enlighten_ the lower stratum, keep clear the atmosphere above, and thus will the colleges give the common schools their clearest support. Take care of the former, and they will take care of the latter," &c., &c. This is hardly better than another argument, employed by the same reasoner in favor of what he calls "practical knowledge." "Our five later Presidents," he says, "were men who were never taught to chop logic _secundum artem_, nor to play shuttlecock with abstractions in college halls." Now it is well known that the four early Presidents who preceded them were not only men of liberal education, but eminent for learning and the highest mental culture. They _had_ learned to deal with abstractions, and to reason _secundum artem_ in college halls. To which side of the scale the real force of this argument inclines, we believe our intelligent readers of all parties may well be trusted to decide.