or in a public school. Narrowness is more easily produced than breadth....

Having in his mind the confusion between teaching and examining which has befogged the whole discussion of the question in England, Mr. E. A. Freeman, the historian, declared against any university teaching of English literature. Mr. Collins quotes Mr. Freeman as writing, "there are many things fit for a man's personal study which are not fit for university examinations. One of these is literature." That literature "cultivates the taste, educates the sympathies, enlarges the mind,” Mr. Freeman makes no attempt to deny; "only we cannot examine in tastes and sympathies," is his reply. Now, if this proves anything, it proves too much. It is an argument, not against teaching English literature only, but against teaching Latin literature and Greek literature. But Mr. Freeman and those who hold with him have not yet suggested that the universities of Oxford and Cambridge should give up the teaching of Greek literature.

There is indeed a difference between the teaching of English literature and the teaching of Greek literature. The texts of the great Greek authors, like the texts of the great English authors, may serve for grammatical instruction and for mere linguistic drill; or they may, the ancient as well as the modern, be used to cultivate the taste, educate the sympathy, and enlarge the mind.

Such exposition differs little from exposition pure and simple. Only, it may be made more forcible, considering that it is the handmaid of argument, that it is intended to clear away error as well as enlighten ignorance, that it deals not only with truth as concreted in isolated facts, but also with larger truths as expressive of complex relations between these facts.

It may be well to begin this exercise with the argument of some such simple questions as those alluded to above. The subject offered at the head of the exercise will entail a somewhat abstruse discussion of the term selfishness.

EXERCISE XLVIII.

INDUCTIVE REASONING.

Subjects:

Is the Love of Money the Root of All Evil?—1 Tim. vi: 10.
Still Waters Run Deep.

Heat Expands and Cold Contracts.

Undue Glorification of Self-made Men.

The Virtues of Cold Water as a Universal Beverage.
Whatever Is, Is Right.

Allusion has been made to the fact that many errors are prevalent which a simple appeal to facts is sufficient to expose. If people examined facts in the first place, or at any rate examined large numbers of facts, before they ventured upon broad general statements, they would be saved from many of these errors. The difficulty in the majority of cases is that the process of inductive reasoning has been too hasty or else there has been no such reasoning at all. Perhaps the appearance of a comet in the heavens is accompanied or followed by some great national or other catastrophe on earth. The thoughtless man does not stop to consider that this may be a mere chance coincidence, but assumes that there must be some vital connection between the two events, and immediately upon the appearance of another comet confidently predicts a similar disaster. The thoughtful man on the contrary is not so ready to assume this connection, but waits to see if the coincidence will be observed a second and a third and a tenth time before he will express even a

provisional opinion. He is the inductive reasoner. He recognizes that one instance is not sufficient to prove the existence of a law; that laws are arrived at only by long observation and careful comparison.

Perhaps on no subject are men so prone to generalize on the strength of a few instances as on the subject of weather, and so we have numberless "weather signs." If the sun shines on a certain day known as "ground-hog day," spring will not open for six weeks. If it rains on Easter Sunday it will rain every Sunday thereafter for seven weeks. "A green Christmas, a white Easter," etc., etc. But the majority of such statements express probabilities only, not laws. Many of them are even counter to probability. Some one has observed them to be true once or twice and taken the rest for demonstrated. To prove their unreliability as general statements we have only to extend the series of observations. A dozen concordant observations do not definitively prove; one discordant one disproves.

Bearing in mind this last truth, it is usually not very difficult to expose an error which has grown out of imperfect induction. It requires only the same appeal to facts upon which we relied in the last exercise but one. With this difference, however: the kind

of error alluded to in that exercise was due to a thoughtless or willful disregard of facts; the kind of error alluded to here has a certain show of truth because it seems to be supported by facts, the only difficulty being that it is supported by too few of them. The refutation of this last may require an acuteness of perception or a patience in investigation not possessed by many, or it may depend on some fortunate discovery

of one invalidating instance among a host of corroborative ones.

Expose if you can any fallacy expressed or implied in the subjects for discussion offered at the head of this exercise.

EXERCISE XLIX.

INDUCTIVE REASONING (CONTINUED).

Subjects:

All Dream Images Derived Solely from Waking Sensations.
Some Relations between Animals and Plants in the Struggle for
Existence.

The kind of argument contemplated in the last exercise was destructive, not constructive. That is, it was devoted to the overthrow of errors that may have arisen from imperfect induction - a matter, we found, often not difficult. The opposite process, like most constructive processes, is not so simple. But let us, if possible, get a clear idea of what induction is, before we attempt to establish any truth by it.

We expose a piece of oak wood to a flame; it catches fire.

We try a piece of hickory, with a similar result. We try ash, maple, pine, mahogany; in every case the same phenomenon results-ignition. We conclude that wood is ignitible. We subject gold, silver, iron, lead, bismuth, platinum, to heat; all melt at some temperature or other. We say, metals are fusible. This is inductive reasoning. Logical induction then is the process of discovering general laws-laws which will be found

true throughout entire classes of particulars. These laws are reached only by carefully examining and comparing large numbers of particular instances.

How can we be sure that because twenty metals are fusible, a twenty-first will be? How can we be sure that the laws arrived at by this inductive process will hold true in cases not yet examined? We can not be sure. And herein lies the difference between perfect and imperfect induction. Where all the similar cases that can possibly exist have been examined, then only is the induction perfect and the truth arrived at eternally secure. It may be unassailably true that every state in the United States has a divorce law; it is by no means so certain that every citizen of the United States advocates a divorce law of some kind. So soon as we resort to imperfect induction we render ourselves liable to error. Not only ignorant weather prophets but great scientists and philosophers often go astray here. For a long time astronomers felt practically certain that all the satellites in our system revolved about their planets in the same direction. But satellites of Uranus and Neptune were discovered which revolved in the opposite direction.

And yet we make use of imperfect induction. The great majority of our so-called general truths are founded upon it. Rarely are all the particular instances within our reach. They lie beyond us, in the future, out in the universe, we know not where. Nevertheless, we venture to make general assertions in regard to them on the strength of the instances within our reach. We do so because we know we may be right, and because we want some anchorage, even though a temporary one, among the shifting sands of doubt. Scientific induction,