This a priori method is most apt of all to provoke real thinking. It should therefore be taken up before any of the others. Not only is it best for making you think deeply, but it will be more likely than any of the others to make you think originally. However, whether attended by great or little success, this method should be followed by others. Not the least fruitful of these would be the evolutionary. This, of course, would consist in studying the history of shorthand, finding out the direction in which it has been tending, and thus anticipating in some degree its future development. As this method is comparative we would naturally be led from it to comparing the shorthand systems of to-day, and assaying the good and bad qualities of each. These could only be assayed if we knew something of shorthand theory, and thus our experience with the deductive or a priori method would be of service. Implied in here is a method of different nature than any we have yet discussed, but one of immense help. In turning from the deductive method to a study of shorthand systems which others have developed, you have an opportunity to compare the results of your own thinking with those obtained by others. If you have failed to solve the question in as good a manner as these others, you can ask yourself wherein and why your own reflections and ingenuity fell short. If you follow this method with all problems—i.e., thinking a thing out for yourself before looking up what others have thought—you will soon improve your thinking surprisingly. The method is capable of application in every problem, from inventing an adding machine to trying to find how the plumber got that $3.46 on the bill. But to return to shorthand. We still have the empirical and experimental methods. In this particular case the difference between them would be simply one of degree. We could find, for instance, what systems were used by the fastest shorthand writers; but we could get nothing conclusive from this, for we would have to make allowance for the natural ability and length of training of these writers. From merely looking at two outlines or characters, it is often difficult to tell which can be written faster. This could only be tested by writing hundreds in a row and finding the time it took to write the same number of each. Of course such experiment is capable of indefinite expansion.