The relation between invention and the engineering arts, and especially between invention and machines, will warrant a short review of the matter here; or even if this reason were wanting, there is a sufficient one in the fact that one of the first aims of an engineering apprentice is to invent something; and as the purpose here is, so far as the limits will permit, to say something upon each subject in which a beginner has an interest, invention must not be passed over.

It has been the object thus far to show that machines, processes, and mechanical manipulation generally may be systematised and generalised to a greater or less extent, and that a failure to reduce mechanical manipulation and machine construction to certain rules and principles can mainly be ascribed to our want of knowledge, and not to any inherent difficulty or condition which prevents such solution. The same proposition is applicable to invention, with the difference that invention, in its true sense, may admit of generalisation more readily than machine processes.

Invention, as applied to mechanical improvements, should not mean chance discovery. Such a meaning is often, if not generally, attached to the term invention, yet it must be seen that results attained by a systematic course of reasoning or experimenting can have nothing to do with chance or even discovery. Such results partake more of the nature of demonstrations, a name peculiarly suitable for such inventions as are the result of methodical purpose.

In such sciences as rest in any degree upon physical experiment, like chemistry, to experiment without some definite object may be a proper kind of research, and may in the future, as it has in the past, lead to great and useful results; but in mechanics the case is different; the demonstration of the conservation of force, and the relation between force and heat, have supplied the last link in a chain of principles which may be said to comprehend all that we are called upon to deal with in dynamical science, and there remains but little hope of developing anything new or useful by discovery alone. The time has been, and has not yet passed away, when even the most unskilled thought their ability to invent improvements in machinery equal with that of an engineer or skilled mechanic; but this is now changed; new schemes are weighed and tested by scientific standards, in many cases as reliable as actual experiments. A veil of mystery which ignorance of the physical sciences had in former times thrown around the mechanic arts, has been cleared away; chance discovery, or mechanical superstition, if the term may be allowed, has nearly disappeared. Many modern engineers regard their improvements in machinery as the exercise of their profession only, and hesitate about asking for protective grants to secure an exclusive use of that which another person might and often does demonstrate, as often as circumstances call for such improvement. There are of course new articles of manufacture to be discovered, and many improvements in machinery which may be proper subject matter for patent rights; improvements which in all chance would not be made for the term of a patent, except by the inventor; but such cases are rare; and it is fair to assume that unless an invention is one which could not have been regularly deduced from existing data, and one that would not in all probability have been made for a long term of years by any other person than the inventor, such an invention cannot in fairness become the property of an individual without infringing the rights of others.

It is not the intention to discuss patent law, nor even to estimate what benefits have in the past, or may in the future, be gained to technical industry by the patent system, but to impress engineering apprentices with a better and more dignified appreciation of their calling than to confound it with chance invention, and thereby destroy that confidence in positive results which has in the past characterised mechanical engineering; also to caution learners against the loss of time and effort too often expended in searching after inventions.

It is well for an apprentice to invent or demonstrate all that he can—the more the better; but as explained in a previous place, what is attempted should be according to some system, and with a proper object. Time spent groping in the dark after something of which no definite conception has been formed, or for any object not to fill an ascertained want, is generally time lost. To demonstrate or invent, one should begin methodically, like a bricklayer builds a wall, as he mortars and sets each brick, so should an engineer qualify, by careful study, each piece or movement that is added to a mechanical structure, so that when done, the result may be useful and enduring.

As remarked, every attempt to generate anything new in machinery should be commenced by ascertaining a want of improvement. When such a want has been ascertained, attention should be directed first to the principles upon which such want or fault is to be remedied. Proper mechanism can then be supplied like the missing links in a chain. Propositions thus stated may fail to convey the meaning intended; this systematic plan of inventing may be better explained by an example.

Presuming the reader to remember what was said of steam hammers in another place, and to be familiar with the uses and general construction of such hammers, let it be supposed steam-hammers, with the ordinary automatic valve action, those that give an elastic or steam-cushioned blow, are well known. Suppose further that by analysing the blows given by hammers of this kind, it is demonstrated that dead blows, such as are given when a hammer comes to a full stop in striking, are more effectual in certain kinds of work, and that steam-hammers would be improved by operating on this dead-stroke principle.

Such a proposition would constitute the first stage of an invention by demonstrating a fault in existing hammers, and a want of certain functions which if added would make an improvement.

Proceeding from these premises, the first thing should be to examine the action of existing valve gear, to determine where this want of the dead-stroke function can best be supplied, and to gain the aid of such suggestions as existing mechanism may offer, also to see how far the appliances in use may become a part of any new arrangement.

By examining automatic hammers it will be found that their valves are connected to the drop by means of links, producing coincident movement of the piston and valve, and that the movement of one is contingent upon and governed by the other. It will also be found that these connections or links are capable of extension, so as to alter the relative position of the piston and valve, thereby regulating the range of the blow, but that the movement of the two is reciprocal or in unison. Reasoning inductively, not discovering or inventing, it may be determined that to secure a stamp blow of a hammer-head, the valve must not open or admit steam beneath the piston until a blow is completed and the hammer has stopped.

At this point will occur one of those mechanical problems which requires what may be called logical solution. The valve must be moved by the drop; there is no other moving mechanism available; the valve and drop must besides be connected, to insure coincident action, yet the valve requires to move when the drop is still. Proceeding inductively, it is clear that a third agent must be introduced, some part moved by the drop, which will in turn move the valve, but this intermediate agent so arranged that it may continue to move after the hammer-drop has stopped.

This assumed, the scheme is complete, so far as the relative movement of the hammer-drop and the valve, but there must be some plan of giving motion to this added mechanism. In many examples there may be seen parts of machinery which continue in motion after the force which propels them has ceased to act; cannon balls are thrown for miles, the impelling force acting for a few feet only; a weaver's shuttle performs nearly its whole flight after the driver has stopped. In the present case, it is therefore evident that an independent or subsequent movement of the valves may be obtained by the momentum of some part set in motion during the descent of the hammer-head.

To sum up, it is supposed to have been determined by inductive reasoning, coupled with some knowledge of mechanics, that a steam hammer, to give a dead blow, requires the following conditions in the valve gearing:

1. That the drop and valve, while they must act relatively, cannot move in the same time, or in direct unison.

2. The connection between the hammer drop and valve cannot be positive, but must be broken during the descent of the drop.

3. The valve must move after the hammer stops.

4. To cause a movement of the valve after the hammer stops there must be an intermediate agent, that will continue to act after the movement of the hammer drop has ceased.

5. The obvious means of attaining this independent movement of the valve gear, is by the momentum of some part set in motion by the hammer-drop, or by the force of gravity reacting on this auxiliary agent.

The invention is now complete, and as the principles are all within the scope of practical mechanism, there is nothing left to do but to devise such mechanical expedients as will carry out the principles laid down. This mechanical scheming is a second, and in some sense an independent part of machine improvement, and should always be subservient to principles; in fact, to separate mechanical scheming from principles, generally constitutes what has been called chance invention.

Referring again to the hammer problem, it will be found by examining the history that the makers of automatic-acting steam-hammers capable of giving the dead stamp blow, have employed the principle which has been described. Instead of employing the momentum, or the gravity of moving parts, to open the valve after the hammer stops, some engineers have depended upon disengaging valve gear by the concussion and jar of the blow, so that the valve gearing, or a portion of it, fell and opened the valve. The 'dead blow gear,' fitted to the earlier Nasmyth, or Wilson, hammers, was constructed on the latter plan, the valve spindle when disengaged being moved by a spring.

I will not consume space to explain the converse of this system of inventing, nor attempt to describe how a chance schemer would proceed to hunt after mechanical expedients to accomplish the valve movement in the example given.

Inventions in machine improvement, no matter what their nature, must of course consist in and conform to certain fixed modes of operating, and no plan of urging the truth of a proposition is so common, even with a chance inventor, as to trace out the 'principles' which govern his discovery.

In studying improvements with a view to practical gain, a learner can have no reasonable hope of accomplishing much in fields already gone over by able engineers, nor in demonstrating anything new in what may be called exhausted subjects, such as steam-engines or water-wheels; he should rather choose new and special subjects, but avoid schemes not in some degree confirmed by existing practice.

It has been already remarked that the boldness of young engineers is very apt to be inversely as their experience, not to say their want of knowledge, and it is only by a strong and determined effort towards conservatism, that a true balance is maintained in judging of new schemes.

The life of George Stephenson proves that notwithstanding the novelty and great importance of his improvements in steam transit, he did not "discover" these improvements. He did not discover that a floating embankment would carry a railway across Chat Moss, neither did he discover that the friction between the wheels of a locomotive and the rails would enable a train to be drawn by tractive power alone. Everything connected with his novel history shows that all of his improvements were founded upon a method of reasoning from principles and generally inductively. To say that he "discovered" our railway system, according to the ordinary construction of the term, would be to detract from his hard and well-earned reputation, and place him among a class of fortunate schemers, who can claim no place in the history of legitimate engineering.

Count Rumford did not by chance develope the philosophy of forces upon which we may say the whole science of dynamics now rests; he set out upon a methodical plan to demonstrate conceptions that were already matured in his mind, and to verify principles which he had assumed by inductive reasoning. The greater part of really good and substantial improvements, such as have performed any considerable part in developing modern mechanical engineering, have come through this course of first dealing with primary principles, instead of groping about blindly after mechanical expedients, and present circumstances point to a time not far distant when chance discovery will quite disappear.