In fact, a very wise strategical direction has as one of its most important functions the division of study and labor among various lines of action, and in deciding which lines are important and which not:  and for this reason may—­and often does—­limit labor, and therefore perfection of result, along lines which a less wise strategy would not limit.  Illustrations of the casting aside of rigid and difficult forms of drill during the past fifty years in armies, and the substitution of more easy methods are numerous.  This does not indicate, however, that a wise strategy may not encourage rigid forms of drill, for the army which is directed with the greatest strategical skill is the German, and no army has more precise methods, not only of procedure, but of drill.  The Prussian army of Frederick William which Frederick the Great inherited was not more rigidly drilled in some particulars than the German army of to-day, fought by Frederick the Great’s great-great-great-grandnephew, William II.

So we see that a wise and far-sighted strategy does not necessarily either frown on or encourage attention to details; it merely regulates it, deciding in each case and for each purpose what degree of attention to detail is best.

The most obvious work of naval strategy, and therefore the work that impresses people most, is in directing naval forces against an enemy in war.  But it is clear that before this can be done effectively strategy must first have made plans of preparation in time of peace; and it is equally clear that, previous to this, strategy must first determine the units of the force and their relation to each other:  it must, in other words, design the machine.

Evidently, therefore, the work of strategy is three-fold:  first, to design the machine; second, to prepare it for war; and, third, to direct its operations during war.

A navy being a machine composed of human and material parts, it is clear that the work of designing it correctly should take account of all the parts at the outset; and not only this—­the whole design should be completed before any parts are made and put together if the best results are to be obtained.  This is the practice in making material machines in manufacturing establishments—­and no other practice there could be successfully pursued.  It is the outcome of the experience of tens of thousands of men for many years—­and the result of the expenditure of tons of money.

This remark as to manufacturing establishments does not include the development of new ideas, for which experimentation or original research is needed; because it is sometimes necessary, when venturing into untrodden fields, to test out by mere trial and error certain parts or features before determining enough of their details to warrant incorporating them in the drawing of the whole machine.  Similarly, some experiments must be made in the methods, organization, and material of the naval machine; but in this, case, as in the case of manufacturing establishments, the experimental work, no matter how promising or alluring, must be recognized as of unproved and doubtful value; and no scheme, plan, or doctrine must be incorporated in the naval machine, or allowed to pose as otherwise than experimental, until successful trials shall have put it beyond the experimental stage.