ON VARIATIONS IN THE VALUE OF THE MONETARY STANDARD. 295 



negative, required in the amount of currency, in order that tlie level of 

 prices may be restored. 



It would he foreign to the spirit of this memorandum to dwell upon ordinary 

 statistical difficulties. But there is one scruple inherent in the nature of the 

 metretic art which even with the progre.ss of statistical technique does not seem 

 likely to he removed. The method under consideration requires the determination 

 of a certain residue, viz., total volume of transactions minus the portion effected 

 by credit, V — in Dr. Krals notation. Now each of these quantities, and 

 (> fortiori their difference, is subject to an error of measurement. And statistics 

 must be much more perfect than there is any prospect of their being in the imme- 

 diate future in order that the error incident to each of these measurables should 

 not exceed a hundredth part of the same. But the bundredtli part of the total 

 transactions is a quantity of about the same order as that which it is sought to 

 determine, namely, the amount of transactions in hard cash. The latter quantity, 

 therefore, will be apt to be lost in a fringe of error. And, though the methods of 

 determining V and C are likely to improve, yet the ratio of V — to V or C is 

 certain to diminish, so that the precariousness of the calculation may well remain 

 constant. 



Upon the whole it seems that in the present state of science we must 

 abandon the sort of realLsm which seeks an additional entity behind the 

 phenomena of varying prices. We must resign the fond idea of finding 

 in the mean variation of price any quantity more objective than itself, any 

 measure of its cause verifiable by an independent statistical investigation. 

 We must be content with measuring the shadows ; the objects behind 

 them are beyond our reach. The cause of the observed phenomenon may 

 be vaguely indicated as the changed relation between shining orb and 

 opaque bodies ; but there is wanting the mathematical science which 

 should express the varying length of shadow as a definite function of the 

 position of the sun. 



The only question is whether we should not adopt a less, not a more, 

 objective quasitum than the type above described ; whether, even where 

 we can use the semi-objective type peculiar to this and the preceding 

 section, it would not be better to use the more subjective formulae inves- 

 tigated in the earlier sections. The present writer, following Laplace, has 

 maintained ' that, even in the case of physical observations relating to a 

 real thing, the proper method of combination is not so much that which 

 is ' most probably ' ^ correct, most frequently in the long run the true 

 measure, but that which may ' most advantageously ' be employed. A 

 fortiori, when our qucesitiim is at best a type, the proper mean may well 

 be not the ratio which is presented by the greatest number of (independ- 

 ently oscillating) prices,^ but thai ratio which in reference to human uses 

 it is best to adopt in any general regulation.'' However a peculiar import- 



' MttreWke, part ii. 

 - Laplace. 



' In the case of our metaphorical shadows suppose that the scope and end of the 

 measurement was to ascertain Whether and by how much shade for the use of man 

 and his cattle was increasing or riecreasing with the change of hour. The determina- 

 tion of a mean variation in the length of shadows would be useful only as a step 

 towards that end. It would be better to aim directly at the end, and combine 

 arithiDetically the length of the shadows multiplied by the corresponding breadth ; 

 this system of weights being now determined, not on the principle proper to this 

 section (see above, p. 290), but on the ground that the broader trees are the more 

 umbrageous. 



* Eead Professor Foxwell's very able lecture on Irregiilarity of EmjJlorjment and 

 Flvctvations of Prices, anA consider wJiat it is, what sort of mean or function of 



