in question after the labor-saving invention was introduced than before. This, of course, would depend upon the degree of inelasticity shown by the demand. But it is pretty certain that more capital will be used, assuming that the invention is a labor-saving rather than a capitalsaving device. There is no reason to suppose that any cheapening of the product would actually reduce the amount consumed. And, if even the same number of units of product are required after the improvement as before, more capital would be used in their production, since one effect of introducing the labor-saving device is to require more capital and less labor in the production of each unit. Here, therefore, would be a case where, from the side of production at least, capital would gain and labor lose, not only relatively, but absolutely.

As already suggested, some inventions are capitalsaving rather than labor-saving. These may be grouped under two classes: first, labor-saving improvements, introduced into the manufacture of machinery and other forms of capital; second, inventions which enable a cheaper kind of machine to displace a more expensive kind. An invention of the first type, tho labor-saving rather than capital-saving in the industry where it is itself employed, becomes capital-saving rather than laborsaving for the industry in which its product is used. Electric cranes and other labor-saving devices in a rolling mill for the manufacture of steel rails will serve as an illustration. In the rolling mill they are labor-saving devices pure and simple; but steel rails are a part of the capital of the railroad, and, in so far as these improvements make rails cheaper, they save capital to the railroad company. They enable the company to construct its road and maintain it at a given standard of efficiency at a lower cost than would otherwise be possible. Since our quantitative notions of capital are always expressed

in terms of value, when there is less value in the rails, there is less capital in the road, other things being equal. That is to say, capital is saved by cheapening the price of steel rails.

An invention of the second type looks at first glance like a capital-saving device, pure and simple. Logically, however, it does not differ so very much from an invention of the first kind. Inventing a new and cheaper machine to do the work formerly done by an expensive machine has about the same effect, so far as our problem is concerned, as inventing a cheaper way of making the older machine. The reason the new machine is cheaper than the older one is probably because it takes less labor to make it. When this is the case, it resembles in all essential particulars the cheapening of the process of making the older machine. In both cases, labor is saved in the industry which supplies productive goods to another industry, but capital is saved in the latter industry. The general result in either case is a saving in both labor and capital to the community as a whole. Whether the saving of labor is greater than that of capital or vice versa is impossible to say off-hand. That question depends upon a great variety of circumstances. Fortunately, it is not necessary to our purpose that this question be answered in any individual case. The most that can be said in advance is that it is extremely unlikely that the margin of difference in favor of the laborer, if there is such a margin at all in these classes of inventions, will be large enough to overbalance the very distinct margin in favor of the capitalists in those classes of inventions which are merely labor-saving and not capital-saving in any sense.

We have still to consider the possible case of an invention which is merely capital-saving and not laborsaving in any sense. This would be an invention which would substitute a cheaper for a dearer form of capital,

and which would effect the cheapening of the new form without any saving of labor in its production. Such a thing appears impossible, and would certainly be very difficult to find in actual fact. But there may be certain cases which look like it. It is possible, for example, that coal might become cheaper, not through any saving in the labor of mining or transporting it, but through the discovery of new sources of supply, thus reducing the monopoly profits which now form a part of the present price. That is to say, the labor cost of producing a ton of coal might be as high as before, but the price might fall, and the reduction in price come out of the profits of monopoly. This cheapening of coal might in turn cheapen the production of certain forms of capital without any reduction in the labor of making them. So far the result is wholly due to a new discovery of natural resources rather than to an invention. But the cheapening of fuel may make practicable a certain type of machine which had formerly been impracticable.

Let us suppose that, in the making of machine A, the labor cost is 50 and the capital cost is 50, of which the fuel cost is one-fifth, or 10, making a total cost of 100: whereas, in the making of machine B, the labor cost is likewise 50, but the capital cost is 60, of which the fuel cost is two-thirds, or 40, making a total cost of 110. A is the cheaper machine, and, assuming that they are equally efficient, A will be used in preference to B. B will not appear in actual use at all, even tho its concept might exist in the mind of an inventor and a model of it might be on exhibition in various places. But a fall of 50 per cent. in the price of fuel would reduce the cost of A to 95 and that of B to 90. B would now be the cheaper machine, and would come into actual use as a substitute for A. Capital would be saved to the factories using these machines, and this saving of capital would not be accomplished by

any saving in labor. This appears, outwardly at least, to be an economy resulting from the substitution of a cheaper for a dearer machine, tho in reality it is directly traceable to the discovery of a new coal field, and therefore should be excluded from consideration.

In fact, if we rule out the influence, direct or indirect, of discoveries of new lands and natural resources, it would be difficult to imagine any mechanical invention which could prove economical without saving labor somewhere. If there are such cases, they must be so rare as to have little influence on the net result of inventions in general. We must expect, therefore, to find that the net result of all inventions is and has been to save labor more than to save capital, or to increase the employment of capital more rapidly than that of labor. This means that, as the result of inventions alone, capital comes to play a larger and larger rôle in industry.

This is capable of being tested statistically, tho, so far as the writer is aware, no such statistical test has yet been applied. In the absence of such a test, we must rely upon general observation. Few observers would deny that more capital is actually used in connection with each unit of labor, say with each laborer, now than before the rise of the present régime of machine industry; nor are many likely to deny that the proportion of capital to labor is still increasing with every advance in mechanical invention. It is doubtful if a single industry can be found in which the amount of capital used by each laborer has diminished in recent times, while the almost universal rule has been an increase of capital out of proportion to the increase of labor.

There can be but one conclusion drawn from these facts, if they are admitted to be facts; namely, that the opportunities for those who are in a position to supply capital have increased more rapidly than the oppor

tunities for those who are in a position to supply only their labor. In other words, the progress of invention has caused a shifting in the relative demand for the services of the different classes in the industrial world. The services of the manual laborers, especially of the lower classes where mere muscular strength is an important element in their productiveness, are becoming of less relative importance, but the services of those who supply capital, as well as of engineers and others who supply certain high types of labor, are becoming relatively more important. Stated negatively, one result of this mass of inventions has been that a manual laborer of the lower grades can be spared with less relative loss than formerly, because his work can now be done by machines, while the loss of a given amount of capital, now that we have so many uses for it, would occasion a relatively greater loss than would have been the case if there had been fewer inventions. Under strictly democratic institutions, the world rewards most highly those whom it most needs, or at least those whom it thinks it most needs. One result of invention is to reduce the need, relatively at least, for muscular energy and for every form of mechanical work which can be reduced to routine. Machines can furnish these elements of production; but the use of machines increases the relative need for the services of those who supply capital,that is, of the capitalists. In somewhat technical language the general result of inventions is to reduce the marginal productivity of the lower grades of labor and to increase that of capital.

The growing demand for capital may possibly, it is true, be supplied by any one who has the foresight to anticipate it. But new capital can only be supplied by waiting, by deferring consumption, by saving something out of present income. This is obviously difficult