There is a further logical discrepancy avoided by resorting to the alleged facts of primitive industry, when there was no capital, for the elements out of which to construct a capital concept, instead of going to the current business situation. In a hedonistic-utilitarian scheme of economic doctrine, such as Mr. Clark's, only physically productive agencies can be admitted as efficient factors in production or as legitimate claimants to a share in distribution. Hence capital, one of the prime factors in production and the central claimant in the current scheme of distribution, must be defined in physical terms and delimited by mechanical distinctions. This is necessary for reasons which appear in the succeeding chapter, on The Measure of Consumers' Wealth. On the same page (38), and elsewhere, it is remarked that "business disasters" destroy capital in part. The destruction in question is a matter of values; that is to say, a lowering of valuation, not in any appreciable degree a destruction of material goods. Taken as a physical aggregate, capital does not appreciably decrease through business disasters, but, taken as a fact of ownership and counted in standard units of value, it decreases; there is a destruction of values and a shifting of ownership, a loss of ownership perhaps; but these are pecuniary phenomena, of an immaterial character, and so do not directly affect the material aggregate of the industrial equipment. Similarly, the discussion (pp. 301-314) of how changes of method, as, _e.g._, labor-saving devices, "liberate capital," and at times "destroy" capital, is intelligible only on the admission that "capital" here is a matter of values owned by investors and is not employed as a synonym for industrial appliances. The appliances in question are neither liberated nor destroyed in the changes contemplated. And it will not do to say that the aggregate of "productive goods" suffers a diminution by a substitution of devices which increases its aggregate productiveness, as is implied, _e.g._, by the passage on page 307,[10] if Mr. Clark's definition of capital is strictly adhered to. This very singular passage (pp. 306-311, under the captions, Hardships entailed on Capitalists by Progress, and the Offset for Capital destroyed by Changes of Method) implies that the aggregate of appliances of production is decreased by a change which increases the aggregate of these articles in that respect (productivity) by virtue of which they are counted in the aggregate. The argument will hold good if "productive goods" are rated by bulk, weight, number, or some such irrelevant test, instead of by their productivity or by their consequent capitalised value. On such a showing it should be proper to say that the polishing of plowshares before they are sent out from the factory diminishes the amount of capital embodied in plowshares by as much as the weight or bulk of the waste material removed from the shares in polishing them.